text
stringlengths 0
1.92k
| source
stringlengths 32
167
|
|---|---|
As Deep Neural Networks (DNNs) grow in size and complexity, they often exceed the memory capacity of a single accelerator, necessitating the sharding of model parameters across multiple accelerators. Pipeline parallelism is a commonly used sharding strategy for training large DNNs. However, current implementations of pipeline parallelism are being unintentionally bottlenecked by the automatic differentiation tools provided by ML frameworks. This paper introduces 2-stage backpropagation (2BP). By splitting the backward propagation step into two separate stages, we can reduce idle compute time. We tested 2BP on various model architectures and pipelining schedules, achieving increases in throughput in all cases. Using 2BP, we were able to achieve a 1.70x increase in throughput compared to traditional methods when training a LLaMa-like transformer with 7 billion parameters across 4 GPUs.
|
https://arxiv.org/abs/2405.18047v1
|
We describe this paper as a Sentimental Journey from Hydrodynamics to
Supergravity. Beltrami equation in three dimensions that plays a key role in
the hydrodynamics of incompressible fluids has an unsuspected relation with
minimal supergravity in seven dimensions. We show that just D=7 supergravity
and no other theory with the same field content but different coefficients in
the lagrangian, admits exact two-brane solutions where Arnold-Beltrami fluxes
in the transverse directions have been switched on. The rich variety of
discrete groups that classify the solutions of Beltrami equation, namely the
eigenfunctions of the *d operator on a three-torus, are by this newly
discovered token injected into the brane world. A new quite extensive playing
ground opens up for supergravity and for its dual gauge theories in three
dimensions, where all classical fields and all quantum composite operators will
be assigned to irreducible representations of discrete crystallographic groups.
|
http://arxiv.org/abs/1504.06802v1
|
This paper presents a simple yet effective approach for the poorly investigated task of global action segmentation, aiming at grouping frames capturing the same action across videos of different activities. Unlike the case of videos depicting all the same activity, the temporal order of actions is not roughly shared among all videos, making the task even more challenging. We propose to use activity labels to learn, in a weakly-supervised fashion, action representations suitable for global action segmentation. For this purpose, we introduce a triadic learning approach for video pairs, to ensure intra-video action discrimination, as well as inter-video and inter-activity action association. For the backbone architecture, we use a Siamese network based on sparse transformers that takes as input video pairs and determine whether they belong to the same activity. The proposed approach is validated on two challenging benchmark datasets: Breakfast and YouTube Instructions, outperforming state-of-the-art methods.
|
https://arxiv.org/abs/2412.12829v1
|
As a generalization of acyclic 2-Calabi-Yau categories, we consider
2-Calabi-Yau categories with a directed cluster-tilting subcategory; we study
their cluster-tilting subcategories and the cluster combinatorics that they
encode. We show that such categories have a cluster structure.
Triangulated 2-Calabi-Yau categories with a directed cluster-tilting
subcategory are closely related to representations of certain semi-hereditary
categories, more specifically to representations of thread quivers. Thread
quivers are a tool to classify and study certain semi-hereditary categories
using both quivers and linearly ordered sets (threads).
We study the case where the thread quiver consists of a single thread (so
that representations of this thread quiver correspond to representations of
some linearly ordered set), and show that, similar to the case of a Dynkin
quiver of type $A$, the cluster-tilting subcategories can be understood via
triangulations of an associated cyclically ordered set.
In this way, we gain insight into the structure of the cluster-tilting
subcategories of 2-Calabi-Yau categories with a directed cluster-tilting
subcategory. As an application, we show that every 2-Calabi-Yau category which
admits a directed cluster-tilting subcategory with countably many isomorphism
classes of indecomposable objects has a cluster-tilting subcategory
$\mathcal{V}$ with the following property: any rigid object in the cluster
category can be reached from $\mathcal{V}$ by finitely many mutations. This
implies that there is a cluster map which is defined on all rigid objects, and
thus that there is a cluster algebra whose cluster variables are exactly given
by the rigid indecomposable objects.
|
http://arxiv.org/abs/1611.03836v1
|
In this paper, we provide a notion of $\infty$-bicategories fibred in $\infty$-bicategories which we call 2-Cartesian fibrations. Our definition is formulated using the language of marked biscaled simplicial sets: Those are scaled simplicial sets equipped with an additional collection of triangles containing the scaled 2-simplices, which we call lean triangles, in addition to a collection of edges containing all degenerate 1-simplices. We prove the existence of a left proper combinatorial simplicial model category whose fibrant objects are precisely the 2-Cartesian fibrations over a chosen scaled simplicial set $S$. Over the terminal scaled simplicial set, this provides a new model structure modeling $\infty$-bicategories, which we show is Quillen equivalent to Lurie's scaled simplicial set model. We conclude by providing a characterization of 2-Cartesian fibrations over an $\infty$-bicategory. This characterization then allows us to identify those 2-Cartesian fibrations arising as the coherent nerve of a fibration of $\operatorname{Set}^+_{\Delta}$-enriched categories, thus showing that our definition recovers the preexisting notions of fibred 2-categories.
|
https://arxiv.org/abs/2106.03606v1
|
In this work, we conclude our study of fibred $\infty$-bicategories by providing a Grothendieck construction in this setting. Given a scaled simplicial set $S$ (which need not be fibrant) we construct a 2-categorical version of Lurie's straightening-unstraightening adjunction, thereby furnishing an equivalence between the $\infty$-bicategory of 2-Cartesian fibrations over $S$ and the $\infty$-bicategory of contravariant functors $S^{\operatorname{op}} \to \mathbb{B}\mathbf{\!}\operatorname{icat}_\infty$ with values in the $\infty$-bicategory of $\infty$-bicategories. We provide a relative nerve construction in the case where the base is a 2-category, and use this to prove a comparison to existing bicategorical Grothendieck constructions.
|
https://arxiv.org/abs/2201.09589v3
|
We produce 2-representations of the positive part of affine quantum enveloping algebras on their finite-dimensional counterparts in type $A_n$. These 2-representations naturally extend the right-multiplication 2-representation of $U_q^+(\mathfrak{sl}_{n+1})$ on itself and are closely related to evaluation morphisms of quantum groups. We expect that our 2-representation exists in all simple types and show that the corresponding 1-representation exists in type $D_4$.
|
https://arxiv.org/abs/2502.08039v2
|
Throughout this paper $G$ is a fixed group, and $k$ is a fixed field. All categories are assumed to be $k$-linear. First we give a systematic way to induce $G$-precoverings by adjoint functors using a 2-categorical machinery, which unifies many similar constructions of $G$-precoverings. Now let $\mathcal{C}$ be a skeletally small category with a $G$-action, $\mathcal{C}/G$ the orbit category of $\mathcal{C}$, $(P, \phi) : \mathcal{C} \rightarrow \mathcal{C}/G$ the canonical $G$-covering, and $\mathrm{mod}\mbox{-} \mathcal{C}$, $\mathrm{mod}\mbox{-} (\mathcal{C}/G)$ the categories of finitely generated modules over $\mathcal{C}, \mathcal{C}/G$, respectively. Then it is well known that there exists a canonical G-precovering $(P., \phi.) : \mathrm{mod}\mbox{-} \mathcal{C} \rightarrow \mathrm{mod}\mbox{-} (\mathcal{C}/G)$. By applying the machinery above to this $(P., \phi.)$, new $G$-precoverings $(\mathrm{mod}\mbox{-} \mathcal{C}) / S \rightarrow (\mathrm{mod}\mbox{-} \mathcal{C}/G)/S'$ are induced between the factor categories or localizations of $\mathrm{mod}\mbox{-} \mathcal{C}$ and $\mathrm{mod}\mbox{-} \mathcal{C}/G$, respectively. This is further applied to the morphism category $\mathrm{H}(\mathrm{mod}\mbox{-} \mathcal{C})$ of $\mathrm{mod}\mbox{-} \mathcal{C}$ to have a $G$-precovering $\mathrm{fp}(\mathcal{K}) \rightarrow \mathrm{fp}(\mathcal{K}')$ between the categories of finitely presented modules over suitable subcategories $\mathcal{K}$ and $\mathcal{K}'$ of $\mathrm{mod}\mbox{-}\mathcal{C}$ and $ \mathrm{mod}\mbox{-} \mathcal{C}/G$, respectively.
|
https://arxiv.org/abs/2402.04680v1
|
In this paper we show that the strict and lax pullbacks of a 2-categorical opfibration along an arbitrary 2-functor are homotopy equivalent. We give two applications. First, we show that the strict fibers of an opfibration model the homotopy fibers. This is a version of Quillen's Theorem B amenable to applications. Second, we compute the $E^2$ page of a homology spectral sequence associated to an opfibration and apply this machinery to a 2-categorical construction of $S^{-1}S$. We show that if $S$ is a symmetric monoidal 2-groupoid with faithful translations then $S^{-1}S$ models the group completion of $S$.
|
https://arxiv.org/abs/2010.11173v2
|
The topic of this thesis is the application of distributive laws between
comonads to the theory of cyclic homology. Explicitly, our main aims are: 1) To
study how the cyclic homology of associative algebras and of Hopf algebras in
the original sense of Connes and Moscovici arises from a distributive law, and
to clarify the role of different notions of bimonad in this generalisation. 2)
To extend the procedure of twisting the cyclic homology of a unital associative
algebra to any duplicial object defined by a distributive law. 3) To study the
universality of Bohm and Stefan's approach to constructing duplicial objects,
which we do in terms of a 2-categorical generalisation of Hochschild
(co)homology. 4) To characterise those categories whose nerve admits a
duplicial structure.
|
http://arxiv.org/abs/1605.08992v1
|
In this paper we present $2$-category theory from the perspective of Gray-categories using the graphical calculus of separated surface diagrams. As an extended example we consider cones and limits of $2$-functors. Then we use the canonical adjunction between $2$-computads and $2$-categories to interpret the comparison structure of lax functors and extend the surface diagram calculus with compositor sheets in order to represent and reason about them.
|
https://arxiv.org/abs/2203.08783v1
|
In this article we analyze the structure of $2$-categories of symmetric
projective bimodules over a finite dimensional algebra with respect to the
action of a finite abelian group. We determine under which condition the
resulting $2$-category is fiat (in the sense of \cite{MM1}) and classify simple
transitive $2$-representations of this $2$-category (under some mild technical
assumption). We also study several classes of examples in detail.
|
http://arxiv.org/abs/1904.05798v1
|
In that paper, we prove that the collection of all FRBSU monoidal categories
and the collection of all crossed modules form a 2 category.
|
http://arxiv.org/abs/1512.06981v1
|
Unitary Ribbon Fusion Categories (URFC) formalize anyonic theories. It has been widely assumed that the same category formalizes a topological quantum computing model. However, in previous work, we addressed and resolved this confusion and demonstrated while the former could be any fusion category, the latter is always a subcategory of Hilb. In this paper, we argue that a categorical formalism that captures and unifies both anyonic theories (the Hardware of quantum computing) and a model of topological quantum computing is a braided (fusion) 2-category. In this 2-category, 0-morphisms describe anyonic types and Hom-categories describe different models of quantum computing. This picture provides an insightful perspective on superselection rules. It presents furthermore a clear distinction between fusion of anyons versus tensor products as defined in linear algebra, between vector spaces of 1-morphisms. The former represents a monoidal product and sum between 0-morphisms and the latter a tensor product and direct sum between 1-morphisms.
|
https://arxiv.org/abs/2505.22171v1
|
Copulas are powerful statistical tools for capturing dependencies across data dimensions. Applying Copulas involves estimating independent marginals, a straightforward task, followed by the much more challenging task of determining a single copulating function, $C$, that links these marginals. For bivariate data, a copula takes the form of a two-increasing function $C: (u,v)\in \mathbb{I}^2 \rightarrow \mathbb{I}$, where $\mathbb{I} = [0, 1]$. This paper proposes 2-Cats, a Neural Network (NN) model that learns two-dimensional Copulas without relying on specific Copula families (e.g., Archimedean). Furthermore, via both theoretical properties of the model and a Lagrangian training approach, we show that 2-Cats meets the desiderata of Copula properties. Moreover, inspired by the literature on Physics-Informed Neural Networks and Sobolev Training, we further extend our training strategy to learn not only the output of a Copula but also its derivatives. Our proposed method exhibits superior performance compared to the state-of-the-art across various datasets while respecting (provably for most and approximately for a single other) properties of C.
|
https://arxiv.org/abs/2309.16391v5
|
To achieve sub-picometer sensitivities in the millihertz band, laser interferometric inertial sensors rely on some form of reduction of the laser frequency noise, typically by locking the laser to a stable frequency reference, such as the narrow-linewidth resonance of an ultra-stable optical cavity or an atomic or molecular transition. In this paper we report on a compact laser frequency stabilization technique based on an unequal-arm Mach-Zehnder interferometer that is sub-nanometer stable at $10\,\mu$Hz, sub-picometer at $0.5\,$mHz, and reaches a noise floor of $7\,\mathrm{fm}/\!\sqrt{\mathrm{Hz}}$ at 1 Hz. The interferometer is used in conjunction with a DC servo to stabilize the frequency of a laser down to a fractional instability below $4 \times 10^{-13}$ at averaging times from 0.1 to 100 seconds. The technique offers a wide operating range, does not rely on complex lock acquisition procedures, and can be readily integrated as part of the optical bench in future gravity missions.
|
https://arxiv.org/abs/2308.11325v2
|
Recent studies have used GAN to transfer expressions between human faces. However, existing models have many flaws: relying on emotion labels, lacking continuous expressions, and failing to capture the expression details. To address these limitations, we propose a novel CycleGAN- and InfoGAN-based network called 2 Cycles Expression Transfer GAN (2CET-GAN), which can learn continuous expression transfer without using emotion labels. The experiment shows our network can generate diverse and high-quality expressions and can generalize to unknown identities. To the best of our knowledge, we are among the first to successfully use an unsupervised approach to disentangle expression representation from identities at the pixel level.
|
https://arxiv.org/abs/2211.11570v1
|
We introduce a new family of finite posets which we call 2-chains. These
first arose in the study of 0-Hecke algebras, but they admit a variety of
different characterisations. We give these characterisations, prove that they
are equivalent and derive some numerical results concerning 2-chains.
|
http://arxiv.org/abs/1809.07574v2
|
We perform a detailed study of perturbations around 2-charge circular fuzz-balls and compare the results with the ones obtained in the case of 'small' BHs. In addition to the photon-sphere modes that govern the prompt ring-down, we find a new branch of long-lived QNMs localised inside the photon-sphere at the (meta)stable minimum of the radial effective potential. The latter are expected to dominate late time signals in the form of 'echoes'. Moreover, contrary to 'small' BHs, some 'static' tidal Love numbers are non-zero and independent of the mass, charges and angular momentum of the fuzz-ball. We rely on the recently established connection between BH or fuzz-ball perturbation theory and quantum Seiberg-Witten curves for N = 2 SYM theories, which in turn are related to Liouville CFT via the AGT correspondence. We test our results against numerical results obtained with Leaver's method of continuous fractions or Breit-Wigner resonance method for direct integration and with the WKB approximation based on geodesic motion. We also exclude rotational super-radiance, due to the absence of an ergo-region, and absorption, due to the absence of a horizon.
|
https://arxiv.org/abs/2212.07504v1
|
Let $d$ be a positive square-free integer. In this paper we shall investigate
the structure of the $2$-class group of the cyclotomic $\mathbb{Z}_2$-extension
of the imaginary biquadratic number field $\mathbb{Q}(\sqrt{d},\sqrt{-1})$.
Furthermore, we deduce the structure of the $2$-class group of cyclotomic
$\mathbb{Z}_2$-extension of $\mathbb{Q}(\sqrt{-d})$.
|
http://arxiv.org/abs/2002.03602v3
|
We expand the theory of 2-classifiers, that are a 2-categorical generalization of subobject classifiers introduced by Weber. The idea is to upgrade monomorphisms to discrete opfibrations. We prove that the conditions of 2-classifier can be checked just on a dense generator. The study of what is classified by a 2-classifier is similarly reduced to a study over the objects that form a dense generator. We then apply our results to the cases of prestacks and stacks, where we can thus look just at the representables. We produce a 2-classifier in prestacks that classifies all discrete opfibrations with small fibres. Finally, we restrict such 2-classifier to a 2-classifier in stacks. This is the main ingredient of a proof that Grothendieck 2-topoi are elementary 2-topoi. Our results also solve a problem posed by Hofmann and Streicher when attempting to lift Grothendieck universes to sheaves.
|
https://arxiv.org/abs/2401.16900v3
|
We show that 2-CLUB is NP-hard for distance to 2-club cluster graphs.
|
http://arxiv.org/abs/1903.05425v1
|
We reduce the dynamics of an ensemble of mean-coupled Stuart-Landau oscillators close to the synchronized solution. In particular, we map the system onto the center manifold of the Benjamin-Feir instability, the bifurcation destabilizing the synchronized oscillation. Using symmetry arguments, we describe the structure of the dynamics on this center manifold up to cubic order, and derive expressions for its parameters. This allows us to investigate phenomena described by the Stuart-Landau ensemble, such as clustering and cluster singularities, in the lower-dimensional center manifold, providing further insights into the symmetry-broken dynamics of coupled oscillators. We show that cluster singularities in the Stuart-Landau ensemble correspond to vanishing quadratic terms in the center manifold dynamics. In addition, they act as organizing centers for the saddle-node bifurcations creating unbalanced cluster states as well for the transverse bifurcations altering the cluster stability. Furthermore, we show that bistability of different solutions with the same cluster-size distribution can only occur when either cluster contains at least $1/3$ of the oscillators, independent of the system parameters.
|
https://arxiv.org/abs/2010.06221v2
|
This paper introduces an innovative approach for handling 2D compound
hypotheses within the Belief Function Theory framework. We propose a
polygon-based generic rep- resentation which relies on polygon clipping
operators. This approach allows us to account in the computational cost for the
precision of the representation independently of the cardinality of the
discernment frame. For the BBA combination and decision making, we propose
efficient algorithms which rely on hashes for fast lookup, and on a topological
ordering of the focal elements within a directed acyclic graph encoding their
interconnections. Additionally, an implementation of the functionalities
proposed in this paper is provided as an open source library. Experimental
results on a pedestrian localization problem are reported. The experiments show
that the solution is accurate and that it fully benefits from the scalability
of the 2D search space granularity provided by our representation.
|
http://arxiv.org/abs/1803.08857v1
|
In this paper we explore relaxations of (Williams) coherent and convex
conditional previsions that form the families of $n$-coherent and $n$-convex
conditional previsions, at the varying of $n$. We investigate which such
previsions are the most general one may reasonably consider, suggesting
(centered) $2$-convex or, if positive homogeneity and conjugacy is needed,
$2$-coherent lower previsions. Basic properties of these previsions are
studied. In particular, we prove that they satisfy the Generalized Bayes Rule
and always have a $2$-convex or, respectively, $2$-coherent natural extension.
The role of these extensions is analogous to that of the natural extension for
coherent lower previsions. On the contrary, $n$-convex and $n$-coherent
previsions with $n\geq 3$ either are convex or coherent themselves or have no
extension of the same type on large enough sets. Among the uncertainty concepts
that can be modelled by $2$-convexity, we discuss generalizations of capacities
and niveloids to a conditional framework and show that the well-known risk
measure Value-at-Risk only guarantees to be centered $2$-convex. In the final
part, we determine the rationality requirements of $2$-convexity and
$2$-coherence from a desirability perspective, emphasising how they weaken
those of (Williams) coherence.
|
http://arxiv.org/abs/1606.06043v1
|
The program of internal type theory seeks to develop the categorical model theory of dependent type theory using the language of dependent type theory itself. In the present work we study internal homotopical type theory by relaxing the notion of a category with families (cwf) to that of a wild, or precoherent higher cwf, and determine coherence conditions that suffice to recover properties expected of models of dependent type theory. The result is a definition of a split 2-coherent wild cwf, which admits as instances both the syntax and the "standard model" given by a universe type. This allows us to give a straightforward internalization of the notion of a 2-coherent reflection of homotopical type theory in itself: namely as a 2-coherent wild cwf morphism from the syntax to the standard model. Our theory also easily specializes to give definitions of "low-dimensional" higher cwfs, and conjecturally includes the container higher model as a further instance.
|
https://arxiv.org/abs/2503.05790v1
|
We characterize noncommutative symmetric Banach spaces for which every
bounded sequence admits either a convergent subsequence, or a $2$-co-lacunary
subsequence. This extends the classical characterization, due to R\"abiger.
|
http://arxiv.org/abs/1909.04258v1
|
The 2-colorable perfect matching problem asks whether a graph can be colored with two colors so that each node has exactly one neighbor with the same color as itself. We prove that this problem is NP-complete, even when restricted to 2-connected 3-regular planar graphs. In 1978, Schaefer proved that this problem is NP-complete in general graphs, and claimed without proof that the same result holds when restricted to 3-regular planar graphs. Thus we fill in the missing proof of this claim, while simultaneously strengthening to 2-connected graphs (which implies existence of a perfect matching). We also prove NP-completeness of $k$-colorable perfect matching, for any fixed $k \geq 2$.
|
https://arxiv.org/abs/2309.09786v1
|
In this paper, we combined two types of partitions and introduced 2-colored Rogers-Ramanujan partitions. By finding some functional equations and using a constructive method, some identities have been found. Some Overpartition identities coincide with our findings. A correspondence between colored partitions and overpartitions is provided.
|
https://arxiv.org/abs/2203.15378v1
|
The 2-girth of a 2-dimensional simplicial complex $X$ is the minimum size of
a non-zero 2-cycle in $H_2(X, \mathbb{Z}/2)$. We consider the maximum possible
girth of a complex with $n$ vertices and $m$ 2-faces. If $m = n^{2 + \alpha}$
for $\alpha < 1/2$, then we show that the 2-girth is at most $4 n^{2 - 2
\alpha}$ and we prove the existence of complexes with 2-girth at least
$c_{\alpha, \epsilon} n^{2 - 2 \alpha - \epsilon}$. On the other hand, if
$\alpha > 1/2$, the 2-girth is at most $C_{\alpha}$. So there is a phase
transition as $\alpha$ passes 1/2.
Our results depend on a new upper bound for the number of combinatorial types
of triangulated surfaces with $v$ vertices and $f$ faces.
|
http://arxiv.org/abs/1509.03871v2
|
A well-known theorem of Whitney states that a 3-connected planar graph admits an essentially unique embedding into the 2-sphere. We prove a 3-dimensional analogue: a simply-connected $2$-complex every link graph of which is 3-connected admits an essentially unique locally flat embedding into the 3-sphere, if it admits one at all. This can be thought of as a generalisation of the 3-dimensional Schoenflies theorem.
|
https://arxiv.org/abs/2109.04085v1
|
In this paper, we study 2-complex symmetric composition operators with the conjugation $J$ on the Hardy space $H^2$. More precisely, we obtain the necessary and sufficient condition for the composition operator $C_\phi$ to be 2-complex symmetric when the symbols $\phi$ is an automorphism of $\mathbb D$. We also characterize the 2-complex symmetric composition operator $C_\phi$ on the Hardy space $H^2$ when $\phi$ is a linear fractional self-map of $\mathbb D$.
|
https://arxiv.org/abs/2110.11184v1
|
A cycle $C$ of length $k$ in graph $G$ is extendable if there is another
cycle $C'$ in $G$ with $V(C) \subset V(C')$ and length $k+1$. A graph is cycle
extendable if every non-Hamiltonian cycle is extendable. In 1990 Hendry
conjectured that any Hamiltonian chordal graph (a Hamiltonian graph with no
induced cycle of length greater than three) is cycle extendable, and this
conjecture has been verified for Hamiltonian chordal graphs which are interval
graphs, planar graphs, and split graphs. We prove that any 2-connected
claw-free chordal graph is cycle extendable.
|
http://arxiv.org/abs/1310.2901v4
|
Let $G$ be a graph. A total dominating set in a graph $G$ is a set $S$ of vertices of $G$ such that every vertex in $G$ is adjacent to a vertex in $S$. Recently, the following question was proposed: "Is it true that every connected cubic graph containing a $3$-cycle has two vertex disjoint total dominating sets?" In this paper, we give a negative answer to this question. Moreover, we prove that if we replace $3$-cycle with $4$-cycle the answer is affirmative. This implies every connected cubic graph containing a diamond (the complete graph of order $4$ minus one edge) as a subgraph can be partitioned into two total dominating sets, a result that was proved in 2017.
|
https://arxiv.org/abs/2308.15114v1
|
A Young diagram $Y$ is called wide if every sub-diagram $Z$ formed by a subset of the rows of $Y$ dominates $Z'$, the conjugate of $Z$. A Young diagram $Y$ is called Latin if its squares can be assigned numbers so that for each $i$, the $i$th row is filled injectively with the numbers $1, \ldots ,a_i$, where $a_i$ is the length of $i$th row of $Y$, and every column is also filled injectively. A conjecture of Chow and Taylor, publicized by Chow, Fan, Goemans, and Vondrak is that a wide Young diagram is Latin. We prove a dual version of the conjecture.
|
https://arxiv.org/abs/2311.17670v2
|
Federated Learning harnesses data from multiple sources to build a single model. While the initial model might belong solely to the actor bringing it to the network for training, determining the ownership of the trained model resulting from Federated Learning remains an open question. In this paper we explore how Blockchains (in particular Ethereum) can be used to determine the evolving ownership of a model trained with Federated Learning. Firstly, we use the step-by-step evaluation metric to assess the relative contributivities of participants in a Federated Learning process. Next, we introduce 2CP, a framework comprising two novel protocols for Blockchained Federated Learning, which both reward contributors with shares in the final model based on their relative contributivity. The Crowdsource Protocol allows an actor to bring a model forward for training, and use their own data to evaluate the contributions made to it. Potential trainers are guaranteed a fair share of the resulting model, even in a trustless setting. The Consortium Protocol gives trainers the same guarantee even when no party owns the initial model and no evaluator is available. We conduct experiments with the MNIST dataset that reveal sound contributivity scores resulting from both Protocols by rewarding larger datasets with greater shares in the model. Our experiments also showed the necessity to pair 2CP with a robust model aggregation mechanism to discard low quality inputs coming from model poisoning attacks.
|
https://arxiv.org/abs/2011.07516v1
|
Let F(X_d) be a smooth Fano variety of lines of a hypersurface X_d of degree
d. In this paper, we prove the Griffiths group Griff_1(F(X_d)) is trivial if
the hypersurface X_d is of 2-Fano type. As a result, we give a positive answer
to a question of Professor Voisin about the first Griffiths groups of Fano
varieties in some cases. Base on this result, we prove that
CH_2(X_d)=\mathbb{Z} for a complex smooth $3$-Fano hypersurface X_d whose Fano
variety of lines is smooth.
|
http://arxiv.org/abs/1512.01721v3
|
We discuss the realization of $2d$ $(0,2)$ gauge theories in terms of branes focusing on Brane Brick Models, which are T-dual to D1-branes probing toric Calabi-Yau 4-folds. These brane setups fully encode the infinite class of $2d$ $(0,2)$ quiver gauge theories on the worldvolume of the D1-branes and substantially streamline their connection to the probed geometries. We review various methods for efficiently generating Brane Brick Models. These algorithms are then used to construct $2d$ $(0,2)$ gauge theories for the cones over all the smooth Fano 3-folds and two infinite families of Sasaki-Einstein 7-manifolds with known metrics. This note is based on the author's talk at the Gauged Linear Sigma Models @ 30 conference at the Simons Center for Geometry and Physics.
|
https://arxiv.org/abs/2402.06993v1
|
We initiate a systematic study of 2d (0,2) quiver gauge theories on the
worldvolume of D1-branes probing singular toric Calabi-Yau 4-folds. We present
an algorithm for efficiently calculating the classical mesonic moduli spaces of
these theories, which correspond to the probed geometries. We also introduce a
systematic procedure for constructing the gauge theories for arbitrary toric
singularities by means of partial resolution, which translates to higgsing in
the field theory. Finally, we introduce Brane Brick Models, a novel class of
brane configurations that consist of D4-branes suspended from an NS5-brane
wrapping a holomorphic surface, tessellating a 3-torus. Brane Brick Models are
the 2d analogues of Brane Tilings and allow a direct connection between
geometry and gauge theory.
|
http://arxiv.org/abs/1506.03818v2
|
Relational information between different types of entities is often modelled by a multilayer network (MLN) -- a network with subnetworks represented by layers. The layers of an MLN can be arranged in different ways in a visual representation, however, the impact of the arrangement on the readability of the network is an open question. Therefore, we studied this impact for several commonly occurring tasks related to MLN analysis. Additionally, layer arrangements with a dimensionality beyond 2D, which are common in this scenario, motivate the use of stereoscopic displays. We ran a human subject study utilising a Virtual Reality headset to evaluate 2D, 2.5D, and 3D layer arrangements. The study employs six analysis tasks that cover the spectrum of an MLN task taxonomy, from path finding and pattern identification to comparisons between and across layers. We found no clear overall winner. However, we explore the task-to-arrangement space and derive empirical-based recommendations on the effective use of 2D, 2.5D, and 3D layer arrangements for MLNs.
|
https://arxiv.org/abs/2307.10674v2
|
More serious works on 2D2C, 2D3C, 2C2Dcw1C3D, 3D3C, rotating turbulence,
thin-layer flows, quasi-static magnetohydrodynamics (QSMHD), and all that are
wanted, but we report timely here some studies on locally and globally
2C2Dcw1C3D flows, with the hope to promote smarter and deeper works.
|
http://arxiv.org/abs/1408.1503v5
|
We propose a two-factor authentication (2FA) mechanism called 2D-2FA to address security and usability issues in existing methods. 2D-2FA has three distinguishing features: First, after a user enters a username and password on a login terminal, a unique $\textit{identifier}$ is displayed to her. She $\textit{inputs}$ the same identifier on her registered 2FA device, which ensures appropriate engagement in the authentication process. Second, a one-time PIN is computed on the device and $\textit{automatically}$ transferred to the server. Thus, the PIN can have very high entropy, making guessing attacks infeasible. Third, the identifier is also incorporated into the PIN computation, which renders $\textit{concurrent attacks}$ ineffective. Third-party services such as push-notification providers and 2FA service providers, do not need to be trusted for the security of the system. The choice of identifiers depends on the device form factor and the context. Users could choose to draw patterns, capture QR codes, etc. We provide a proof of concept implementation, and evaluate performance, accuracy, and usability of the system. We show that the system offers a lower error rate (about half) and better efficiency (2-3 times faster) compared to the commonly used PIN-2FA. Our study indicates a high level of usability with a SUS of 75, and a high perception of efficiency, security, accuracy, and adoptability.
|
https://arxiv.org/abs/2110.15872v1
|
Despite recent advances in facial recognition, there remains a fundamental issue concerning degradations in performance due to substantial perspective (pose) differences between enrollment and query (probe) imagery. Therefore, we propose a novel domain adaptive framework to facilitate improved performances across large discrepancies in pose by enabling image-based (2D) representations to infer properties of inherently pose invariant point cloud (3D) representations. Specifically, our proposed framework achieves better pose invariance by using (1) a shared (joint) attention mapping to emphasize common patterns that are most correlated between 2D facial images and 3D facial data and (2) a joint entropy regularizing loss to promote better consistency$\unicode{x2014}$enhancing correlations among the intersecting 2D and 3D representations$\unicode{x2014}$by leveraging both attention maps. This framework is evaluated on FaceScape and ARL-VTF datasets, where it outperforms competitive methods by achieving profile (90$\unicode{x00b0}$$\unicode{x002b}$) TAR @ 1$\unicode{x0025}$ FAR improvements of at least 7.1$\unicode{x0025}$ and 1.57$\unicode{x0025}$, respectively.
|
https://arxiv.org/abs/2505.09073v1
|
Doping of silicon via phosphene exposures alternating with molecular beam
epitaxy overgrowth is a path to Si:P substrates for conventional
microelectronics and quantum information technologies. The technique also
provides a new and well-controlled material for systematic studies of
two-dimensional lattices with a half-filled band. We show here that for a dense
($n_s=2.8\times 10^{14}$\,cm$^{-2}$) disordered two-dimensional array of P
atoms, the full field angle-dependent magnetostransport is remarkably well
described by classic weak localization theory with no corrections due to
interaction effects. The two- to three-dimensional cross-over seen upon warming
can also be interpreted using scaling concepts, developed for anistropic
three-dimensional materials, which work remarkably except when the applied
fields are nearly parallel to the conducting planes.
|
http://arxiv.org/abs/1802.05208v2
|
Deformable registration of two-dimensional/three-dimensional (2D/3D) images of abdominal organs is a complicated task because the abdominal organs deform significantly and their contours are not detected in two-dimensional X-ray images. We propose a supervised deep learning framework that achieves 2D/3D deformable image registration between 3D volumes and single-viewpoint 2D projected images. The proposed method learns the translation from the target 2D projection images and the initial 3D volume to 3D displacement fields. In experiments, we registered 3D-computed tomography (CT) volumes to digitally reconstructed radiographs generated from abdominal 4D-CT volumes. For validation, we used 4D-CT volumes of 35 cases and confirmed that the 3D-CT volumes reflecting the nonlinear and local respiratory organ displacement were reconstructed. The proposed method demonstrate the compatible performance to the conventional methods with a dice similarity coefficient of 91.6 \% for the liver region and 85.9 \% for the stomach region, while estimating a significantly more accurate CT values.
|
https://arxiv.org/abs/2212.05445v1
|
Recent developments in the registration of histology and micro-computed tomography ({\mu}CT) have broadened the perspective of pathological applications such as virtual histology based on {\mu}CT. This topic remains challenging because of the low image quality of soft tissue CT. Additionally, soft tissue samples usually deform during the histology slide preparation, making it difficult to correlate the structures between histology slide and {\mu}CT. In this work, we propose a novel 2D-3D multi-modal deformable image registration method. The method uses a machine learning (ML) based initialization followed by the registration. The registration is finalized by an analytical out-of-plane deformation refinement. The method is evaluated on datasets acquired from tonsil and tumor tissues. {\mu}CTs of both phase-contrast and conventional absorption modalities are investigated. The registration results from the proposed method are compared with those from intensity- and keypoint-based methods. The comparison is conducted using both visual and fiducial-based evaluations. The proposed method demonstrates superior performance compared to the other two methods.
|
https://arxiv.org/abs/2410.14343v1
|
In 2D+3D facial expression recognition (FER), existing methods generate multi-view geometry maps to enhance the depth feature representation. However, this may introduce false estimations due to local plane fitting from incomplete point clouds. In this paper, we propose a novel Map Generation technique from the viewpoint of information theory, to boost the slight 3D expression differences from strong personality variations. First, we examine the HDR depth data to extract the discriminative dynamic range $r_{dis}$, and maximize the entropy of $r_{dis}$ to a global optimum. Then, to prevent the large deformation caused by over-enhancement, we introduce a depth distortion constraint and reduce the complexity from $O(KN^2)$ to $O(KN\tau)$. Furthermore, the constrained optimization is modeled as a $K$-edges maximum weight path problem in a directed acyclic graph, and we solve it efficiently via dynamic programming. Finally, we also design an efficient Facial Attention structure to automatically locate subtle discriminative facial parts for multi-scale learning, and train it with a proposed loss function $\mathcal{L}_{FA}$ without any facial landmarks. Experimental results on different datasets show that the proposed method is effective and outperforms the state-of-the-art 2D+3D FER methods in both FER accuracy and the output entropy of the generated maps.
|
https://arxiv.org/abs/2011.08333v1
|
In this paper, a novel approach via embedded tensor manifold regularization for 2D+3D facial expression recognition (FERETMR) is proposed. Firstly, 3D tensors are constructed from 2D face images and 3D face shape models to keep the structural information and correlations. To maintain the local structure (geometric information) of 3D tensor samples in the low-dimensional tensors space during the dimensionality reduction, the $\ell_0$-norm of the core tensors and a tensor manifold regularization scheme embedded on core tensors are adopted via a low-rank truncated Tucker decomposition on the generated tensors. As a result, the obtained factor matrices will be used for facial expression classification prediction. To make the resulting tensor optimization more tractable, $\ell_1$-norm surrogate is employed to relax $\ell_0$-norm and hence the resulting tensor optimization problem has a nonsmooth objective function due to the $\ell_1$-norm and orthogonal constraints from the orthogonal Tucker decomposition. To efficiently tackle this tensor optimization problem, we establish the first-order optimality condition in terms of stationary points, and then design a block coordinate descent (BCD) algorithm with convergence analysis and the computational complexity. Numerical results on BU-3DFE database and Bosphorus databases demonstrate the effectiveness of our proposed approach.
|
https://arxiv.org/abs/2201.12506v1
|
In this paper, we develop a 2D and 3D segmentation pipelines for fully
automated cardiac MR image segmentation using Deep Convolutional Neural
Networks (CNN). Our models are trained end-to-end from scratch using the ACD
Challenge 2017 dataset comprising of 100 studies, each containing Cardiac MR
images in End Diastole and End Systole phase. We show that both our
segmentation models achieve near state-of-the-art performance scores in terms
of distance metrics and have convincing accuracy in terms of clinical
parameters. A comparative analysis is provided by introducing a novel dice loss
function and its combination with cross entropy loss. By exploring different
network structures and comprehensive experiments, we discuss several key
insights to obtain optimal model performance, which also is central to the
theme of this challenge.
|
http://arxiv.org/abs/1707.09813v1
|
Multi-modal fusion has been proved to help enhance the performance of scene classification tasks. This paper presents a 2D-3D Fusion stage that combines 3D Geometric Features with 2D Texture Features obtained by 2D Convolutional Neural Networks. To get a robust 3D Geometric embedding, a network that uses two novel layers is proposed. The first layer, Multi-Neighbourhood Graph Convolution, aims to learn a more robust geometric descriptor of the scene combining two different neighbourhoods: one in the Euclidean space and the other in the Feature space. The second proposed layer, Nearest Voxel Pooling, improves the performance of the well-known Voxel Pooling. Experimental results, using NYU-Depth-V2 and SUN RGB-D datasets, show that the proposed method outperforms the current state-of-the-art in RGB-D indoor scene classification task.
|
https://arxiv.org/abs/2009.11154v3
|
We present a Multimodal Interlaced Transformer (MIT) that jointly considers 2D and 3D data for weakly supervised point cloud segmentation. Research studies have shown that 2D and 3D features are complementary for point cloud segmentation. However, existing methods require extra 2D annotations to achieve 2D-3D information fusion. Considering the high annotation cost of point clouds, effective 2D and 3D feature fusion based on weakly supervised learning is in great demand. To this end, we propose a transformer model with two encoders and one decoder for weakly supervised point cloud segmentation using only scene-level class tags. Specifically, the two encoders compute the self-attended features for 3D point clouds and 2D multi-view images, respectively. The decoder implements interlaced 2D-3D cross-attention and carries out implicit 2D and 3D feature fusion. We alternately switch the roles of queries and key-value pairs in the decoder layers. It turns out that the 2D and 3D features are iteratively enriched by each other. Experiments show that it performs favorably against existing weakly supervised point cloud segmentation methods by a large margin on the S3DIS and ScanNet benchmarks. The project page will be available at https://jimmy15923.github.io/mit_web/.
|
https://arxiv.org/abs/2310.12817v2
|
Large-scale point cloud generated from 3D sensors is more accurate than its
image-based counterpart. However, it is seldom used in visual pose estimation
due to the difficulty in obtaining 2D-3D image to point cloud correspondences.
In this paper, we propose the 2D3D-MatchNet - an end-to-end deep network
architecture to jointly learn the descriptors for 2D and 3D keypoint from image
and point cloud, respectively. As a result, we are able to directly match and
establish 2D-3D correspondences from the query image and 3D point cloud
reference map for visual pose estimation. We create our Oxford 2D-3D Patches
dataset from the Oxford Robotcar dataset with the ground truth camera poses and
2D-3D image to point cloud correspondences for training and testing the deep
network. Experimental results verify the feasibility of our approach.
|
http://arxiv.org/abs/1904.09742v1
|
The commonly adopted detect-then-match approach to registration finds difficulties in the cross-modality cases due to the incompatible keypoint detection and inconsistent feature description. We propose, 2D3D-MATR, a detection-free method for accurate and robust registration between images and point clouds. Our method adopts a coarse-to-fine pipeline where it first computes coarse correspondences between downsampled patches of the input image and the point cloud and then extends them to form dense correspondences between pixels and points within the patch region. The coarse-level patch matching is based on transformer which jointly learns global contextual constraints with self-attention and cross-modality correlations with cross-attention. To resolve the scale ambiguity in patch matching, we construct a multi-scale pyramid for each image patch and learn to find for each point patch the best matching image patch at a proper resolution level. Extensive experiments on two public benchmarks demonstrate that 2D3D-MATR outperforms the previous state-of-the-art P2-Net by around $20$ percentage points on inlier ratio and over $10$ points on registration recall. Our code and models are available at https://github.com/minhaolee/2D3DMATR.
|
https://arxiv.org/abs/2308.05667v2
|
We presented a 2D/3D MV image registration method based on a Convolutional
Neural Network. Most of the traditional image registration method
intensity-based, which use optimization algorithms to maximize the similarity
between to images. Although these methods can achieve good results for
kilovoltage images, the same does not occur for megavoltage images due to the
lower image quality. Also, these methods most of the times do not present a
good capture range. To deal with this problem, we propose the use of
Convolutional Neural Network. The experiments were performed using a dataset of
50 brain images. The results showed to be promising compared to traditional
image registration methods.
|
http://arxiv.org/abs/1811.11816v1
|
This study considers the 3D human pose estimation problem in a single RGB
image by proposing a conditional random field (CRF) model over 2D poses, in
which the 3D pose is obtained as a byproduct of the inference process. The
unary term of the proposed CRF model is defined based on a powerful heat-map
regression network, which has been proposed for 2D human pose estimation. This
study also presents a regression network for lifting the 2D pose to 3D pose and
proposes the prior term based on the consistency between the estimated 3D pose
and the 2D pose. To obtain the approximate solution of the proposed CRF model,
the N-best strategy is adopted. The proposed inference algorithm can be viewed
as sequential processes of bottom-up generation of 2D and 3D pose proposals
from the input 2D image based on deep networks and top-down verification of
such proposals by checking their consistencies. To evaluate the proposed
method, we use two large-scale datasets: Human3.6M and HumanEva. Experimental
results show that the proposed method achieves the state-of-the-art 3D human
pose estimation performance.
|
http://arxiv.org/abs/1704.03986v2
|
Action recognition and human pose estimation are closely related but both
problems are generally handled as distinct tasks in the literature. In this
work, we propose a multitask framework for jointly 2D and 3D pose estimation
from still images and human action recognition from video sequences. We show
that a single architecture can be used to solve the two problems in an
efficient way and still achieves state-of-the-art results. Additionally, we
demonstrate that optimization from end-to-end leads to significantly higher
accuracy than separated learning. The proposed architecture can be trained with
data from different categories simultaneously in a seamlessly way. The reported
results on four datasets (MPII, Human3.6M, Penn Action and NTU) demonstrate the
effectiveness of our method on the targeted tasks.
|
http://arxiv.org/abs/1802.09232v2
|
Camera localization in 3D LiDAR maps has gained increasing attention due to its promising ability to handle complex scenarios, surpassing the limitations of visual-only localization methods. However, existing methods mostly focus on addressing the cross-modal gaps, estimating camera poses frame by frame without considering the relationship between adjacent frames, which makes the pose tracking unstable. To alleviate this, we propose to couple the 2D-3D correspondences between adjacent frames using the 2D-2D feature matching, establishing the multi-view geometrical constraints for simultaneously estimating multiple camera poses. Specifically, we propose a new 2D-3D pose tracking framework, which consists: a front-end hybrid flow estimation network for consecutive frames and a back-end pose optimization module. We further design a cross-modal consistency-based loss to incorporate the multi-view constraints during the training and inference process. We evaluate our proposed framework on the KITTI and Argoverse datasets. Experimental results demonstrate its superior performance compared to existing frame-by-frame 2D-3D pose tracking methods and state-of-the-art vision-only pose tracking algorithms. More online pose tracking videos are available at \url{https://youtu.be/yfBRdg7gw5M}
|
https://arxiv.org/abs/2309.11335v1
|
We study an active Brownian run-and-tumble particle (ABRTP) model, that consists of an active Brownian run state during which the active velocity of the particle diffuses on the unit circle, and a tumble state during which the active velocity is zero, both with exponentially distributed time. Additionally we add a harmonic trap as an external potential. In the appropriate limits the ABRTP model reduces either to the active Brownian particle model, or the run-and-tumble particle model. Using the method of direct integration the equation of motion, pioneered by Kac, we obtain exact moments for the Laplace transform of the time dependent ABRTP, in the presence or absence of a harmonic trap. In addition we estimate the distribution moments with the help of the Chebyshev polynomials. Our results are in excellent agreement with the experiments.
|
https://arxiv.org/abs/2504.20352v1
|
In this work, we have employed Monte Carlo calculations to study the Ising model on a 2D additive small-world network with long-range interactions depending on the geometric distance between interacting sites. The network is initially defined by a regular square lattice and with probability $p$ each site is tested for the possibility of creating a long-range interaction with any other site that has not yet received one. Here, we used the specific case where $p=1$, meaning that every site in the network has one long-range interaction in addition to the short-range interactions of the regular lattice. These long-range interactions depend on a power-law form, $J_{ij}=r_{ij}^{-\alpha}$, with the geometric distance $r_{ij}$ between connected sites $i$ and $j$. In current two-dimensional model, we found that mean-field critical behavior is observed only at $\alpha=0$. As $\alpha$ increases, the network size influences the phase transition point of the system, i.e., indicating a crossover behavior. However, given the two-dimensional system, we found the critical behavior of the short-range interaction at $\alpha\approx2$. Thus, the limitation in the number of long-range interactions compared to the globally coupled model, as well as the form of the decay of these interactions, prevented us from finding a regime with finite phase transition points and continuously varying critical exponents in $0<\alpha<2$.
|
https://arxiv.org/abs/2409.02033v1
|
As transistor footprint scales down to sub-10 nm regime, the process development for advancing to further technology nodes has encountered slowdowns. Achieving greater functionality within a single chip requires concurrent development at the device, circuit, and system levels. Reconfigurable transistors possess the capability to transform into both n-type and p-type transistors dynamically during operation. This transistor-level reconfigurability enables field-programmable logic circuits with fewer components compared to conventional circuits. However, the reconfigurability requires additional polarity control gates in the transistor and potentially impairs the gain from a smaller footprint. In this paper, vertical transistors with ambipolar MoTe2 channels are fabricated using the transfer-metal method. The efficient asymmetric electrostatic gating in source and drain contacts gives rise to different Schottky barriers at the two contacts. Consequently, the ambipolar conduction is reduced to unipolar conduction due to different Schottky barrier widths for electrons and holes. The current flow direction determines the preferred carrier type. Temperature-dependent measurements reveal the Schottky barrier-controlled conduction in the vertical transistors and confirm different Schottky barrier widths with and without electrostatic gating. Without the complexity overhead from polarity control gates, control-free vertical reconfigurable transistors promise higher logic density and lower cost in future integrated circuits.
|
https://arxiv.org/abs/2309.08746v1
|
Amodal instance segmentation aims to predict the complete mask of the occluded instance, including both visible and invisible regions. Existing 2D AIS methods learn and predict the complete silhouettes of target instances in 2D space. However, masks in 2D space are only some observations and samples from the 3D model in different viewpoints and thus can not represent the real complete physical shape of the instances. With the 2D masks learned, 2D amodal methods are hard to generalize to new viewpoints not included in the training dataset. To tackle these problems, we are motivated by observations that (1) a 2D amodal mask is the projection of a 3D complete model, and (2) the 3D complete model can be recovered and reconstructed from the occluded 2D object instances. This paper builds a bridge to link the 2D occluded instances with the 3D complete models by 3D reconstruction and utilizes 3D shape prior for 2D AIS. To deal with the diversity of 3D shapes, our method is pretrained on large 3D reconstruction datasets for high-quality results. And we adopt the unsupervised 3D reconstruction method to avoid relying on 3D annotations. In this approach, our method can reconstruct 3D models from occluded 2D object instances and generalize to new unseen 2D viewpoints of the 3D object. Experiments demonstrate that our method outperforms all existing 2D AIS methods.
|
https://link.springer.com/chapter/10.1007/978-3-031-19818-2_10
|
Since the appearance of Covid-19 in late 2019, Covid-19 has become an active research topic for the artificial intelligence (AI) community. One of the most interesting AI topics is Covid-19 analysis of medical imaging. CT-scan imaging is the most informative tool about this disease. This work is part of the 3nd COV19D competition for Covid-19 Severity Prediction. In order to deal with the big gap between the validation and test results that were shown in the previous version of this competition, we proposed to combine the prediction of 2D and 3D CNN predictions. For the 2D CNN approach, we propose 2B-InceptResnet architecture which consists of two paths for segmented lungs and infection of all slices of the input CT-scan, respectively. Each path consists of ConvLayer and Inception-ResNet pretrained model on ImageNet. For the 3D CNN approach, we propose hybrid-DeCoVNet architecture which consists of four blocks: Stem, four 3D-ResNet layers, Classification Head and Decision layer. Our proposed approaches outperformed the baseline approach in the validation data of the 3nd COV19D competition for Covid-19 Severity Prediction by 36%.
|
https://arxiv.org/abs/2303.08740v1
|
This paper presents new mappings of 2D and 3D geometrical transformation on
the MorphoSys (M1) reconfigurable computing (RC) prototype [2]. This improves
the system performance as a graphics accelerator [1-5]. Three algorithms are
mapped including two for calculating 2D transformations, and one for 3D
transformations. The results presented indicate an improved performance. The
speedup achieved is explained as well as the advantages in the mapping of the
application. The transformations on an 8x8 RC array were run, and numerical
examples were simulated to validate our results, using the MorphoSys mULATE
program, which simulates MorphoSys operations. Comparisons with other systems
are presented, namely, with Intel processing systems and Celoxica RC-1000 FPGA.
|
http://arxiv.org/abs/1904.12609v1
|
The following convective Brinkman-Forchheimer (CBF) equations (or damped Navier-Stokes equations) with potential \begin{equation*} \frac{\partial \boldsymbol{y}}{\partial t}-\mu \Delta\boldsymbol{y}+(\boldsymbol{y}\cdot\nabla)\boldsymbol{y}+\alpha\boldsymbol{y}+\beta|\boldsymbol{y}|^{r-1}\boldsymbol{y}+\nabla p+\Psi(\boldsymbol{y})\ni\boldsymbol{g},\ \nabla\cdot\boldsymbol{y}=0, \end{equation*} in a $d$-dimensional torus is considered in this work, where $d\in\{2,3\}$, $\mu,\alpha,\beta>0$ and $r\in[1,\infty)$. For $d=2$ with $r\in[1,\infty)$ and $d=3$ with $r\in[3,\infty)$ ($2\beta\mu\geq 1$ for $d=r=3$), we establish the existence of \textsf{\emph{a unique global strong solution}} for the above multi-valued problem with the help of the \textsf{\emph{abstract theory of $m$-accretive operators}}. %for nonlinear differential equations of accretive type in Banach spaces. Moreover, we demonstrate that the same results hold \textsf{\emph{local in time}} for the case $d=3$ with $r\in[1,3)$ and $d=r=3$ with $2\beta\mu<1$. We explored the $m$-accretivity of the nonlinear as well as multi-valued operators, Yosida approximations and their properties, and several higher order energy estimates in the proofs. For $r\in[1,3]$, we {quantize (modify)} the Navier-Stokes nonlinearity $(\boldsymbol{y}\cdot\nabla)\boldsymbol{y}$ to establish the existence and uniqueness results, while for $r\in[3,\infty)$ ($2\beta\mu\geq1$ for $r=3$), we handle the Navier-Stokes nonlinearity by the nonlinear damping term $\beta|\boldsymbol{y}|^{r-1}\boldsymbol{y}$. Finally, we discuss the applications of the above developed theory in feedback control problems like flow invariance, time optimal control and stabilization.
|
https://arxiv.org/abs/2301.01527v2
|
Objective: Radiomics, an emerging tool for medical image analysis, is potential towards precisely characterizing gastric cancer (GC). Whether using one-slice 2D annotation or whole-volume 3D annotation remains a long-time debate, especially for heterogeneous GC. We comprehensively compared 2D and 3D radiomic features' representation and discrimination capacity regarding GC, via three tasks. Methods: Four-center 539 GC patients were retrospectively enrolled and divided into the training and validation cohorts. From 2D or 3D regions of interest (ROIs) annotated by radiologists, radiomic features were extracted respectively. Feature selection and model construction procedures were customed for each combination of two modalities (2D or 3D) and three tasks. Subsequently, six machine learning models (Model_2D^LNM, Model_3D^LNM; Model_2D^LVI, Model_3D^LVI; Model_2D^pT, Model_3D^pT) were derived and evaluated to reflect modalities' performances in characterizing GC. Furthermore, we performed an auxiliary experiment to assess modalities' performances when resampling spacing is different. Results: Regarding three tasks, the yielded areas under the curve (AUCs) were: Model_2D^LNM's 0.712 (95% confidence interval, 0.613-0.811), Model_3D^LNM's 0.680 (0.584-0.775); Model_2D^LVI's 0.677 (0.595-0.761), Model_3D^LVI's 0.615 (0.528-0.703); Model_2D^pT's 0.840 (0.779-0.901), Model_3D^pT's 0.813 (0.747-0.879). Moreover, the auxiliary experiment indicated that Models_2D are statistically more advantageous than Models3D with different resampling spacings. Conclusion: Models constructed with 2D radiomic features revealed comparable performances with those constructed with 3D features in characterizing GC. Significance: Our work indicated that time-saving 2D annotation would be the better choice in GC, and provided a related reference to further radiomics-based researches.
|
https://arxiv.org/abs/2210.16640v1
|
Parkinson's Disease (PD) diagnosis remains challenging. This study applies Convolutional Kolmogorov-Arnold Networks (ConvKANs), integrating learnable spline-based activation functions into convolutional layers, for PD classification using structural MRI. The first 3D implementation of ConvKANs for medical imaging is presented, comparing their performance to Convolutional Neural Networks (CNNs) and Graph Convolutional Networks (GCNs) across three open-source datasets. Isolated analyses assessed performance within individual datasets, using cross-validation techniques. Holdout analyses evaluated cross-dataset generalizability by training models on two datasets and testing on the third, mirroring real-world clinical scenarios. In isolated analyses, 2D ConvKANs achieved the highest AUC of 0.99 (95% CI: 0.98-0.99) on the PPMI dataset, outperforming 2D CNNs (AUC: 0.97, p = 0.0092). 3D models showed promise, with 3D CNN and 3D ConvKAN reaching an AUC of 0.85 on PPMI. In holdout analyses, 3D ConvKAN demonstrated superior generalization, achieving an AUC of 0.85 on early-stage PD data. GCNs underperformed in 2D but improved in 3D implementations. These findings highlight ConvKANs' potential for PD detection, emphasize the importance of 3D analysis in capturing subtle brain changes, and underscore cross-dataset generalization challenges. This study advances AI-assisted PD diagnosis using structural MRI and emphasizes the need for larger-scale validation.
|
https://arxiv.org/abs/2407.17380v2
|
We present multidimensional modeling of convection and oscillations in
main-sequence stars somewhat more massive than the Sun, using three separate
approaches: 1) Using the 3-D planar StellarBox radiation hydrodynamics code to
model the envelope convection zone and part of the radiative zone. Our goals
are to examine the interaction of stellar pulsations with turbulent convection
in the envelope, excitation of acoustic modes, and the role of convective
overshooting; 2) Applying the spherical 3-D MHD ASH (Anelastic Spherical
Harmonics) code to simulate the core convection and radiative zone. Our goal is
to determine whether core convection can excite low-frequency gravity modes,
and thereby explain the presence of low frequencies for some hybrid gamma
Doradus/delta Scuti variables for which the envelope convection zone is too
shallow for the convective blocking mechanism to drive gravity modes; 3)
Applying the ROTORC 2-D stellar evolution and dynamics code to calculate
evolution with a variety of initial rotation rates and extents of core
convective overshooting. The nonradial adiabatic pulsation frequencies of these
nonspherical models are calculated using the 2-D pulsation code NRO. We present
new insights into pulsations for stars of one to two solar masses gained by
multidimensional modeling.
|
http://arxiv.org/abs/1605.04455v1
|
Collagen fiber orientations in bones, visible with Second Harmonic Generation (SHG) microscopy, represent the inner structure and its alteration due to influences like cancer. While analyses of these orientations are valuable for medical research, it is not feasible to analyze the needed large amounts of local orientations manually. Since we have uncertain borders for these local orientations only rough regions can be segmented instead of a pixel-wise segmentation. We analyze the effect of these uncertain borders on human performance by a user study. Furthermore, we compare a variety of 2D and 3D methods such as classical approaches like Fourier analysis with state-of-the-art deep neural networks for the classification of local fiber orientations. We present a general way to use pretrained 2D weights in 3D neural networks, such as Inception-ResNet-3D a 3D extension of Inception-ResNet-v2. In a 10 fold cross-validation our two stage segmentation based on Inception-ResNet-3D and transferred 2D ImageNet weights achieves a human comparable accuracy.
|
https://arxiv.org/abs/1907.12868v1
|
Recently, it was shown that quantum spin Hall insulator (QSHI) phase with a
gap wide enough for practical applications can be realized in the ultra thin
films constructed from two inversely stacked structural elements of trivial
band insulator BiTeI. Here, we study the edge states in the free-standing
Bi$_2$Te$_2$I$_2$ sextuple layer (SL) and the electronic structure of the
Bi$_2$Te$_2$I$_2$ SL on the natural BiTeI substrate. We show that the
topological properties of the Bi$_2$Te$_2$I$_2$ SL on this substrate keep
$\mathbb Z_2$ invariant. We also demonstrate that ultra thin centrosymmetric
films constructed in the similar manner but from related material BiTeBr are
trivial band insulators up to five-SL film thickness. In contrast to
Bi$_2$Te$_2$I$_2$ for which the stacking of nontrivial SLs in 3D limit gives a
strong topological insulator (TI) phase, strong TI is realized in 3D
Bi$_2$Te$_2$Br$_2$ in spite of the SL is trivial. For the last material of the
BiTe$X$ ($X$=I,Br,Cl) series, BiTeCl, both 2D and 3D centrosymmetric phases are
characterized by topologically trivial band structure.
|
http://arxiv.org/abs/1706.08127v1
|
The detection of vascular structures from noisy images is a fundamental
process for extracting meaningful information in many applications. Most
well-known vascular enhancing techniques often rely on Hessian-based filters.
This paper investigates the feasibility and deficiencies of detecting
curve-like structures using a Hessian matrix. The main contribution is a novel
enhancement function, which overcomes the deficiencies of established methods.
Our approach has been evaluated quantitatively and qualitatively using
synthetic examples and a wide range of real 2D and 3D biomedical images.
Compared with other existing approaches, the experimental results prove that
our proposed approach achieves high-quality curvilinear structure enhancement.
|
http://arxiv.org/abs/1902.00550v1
|
We challenge two foundational principles of localization physics by analyzing conductance fluctuations in two dimensions with unprecedented precision: (i) the Thouless criterion, which defines localization as insensitivity to boundary conditions, and (ii) that symmetry determines the universality class of Anderson localization. We reveal that the fluctuations of the conductance logarithm fall into distinct sub-universality classes inherited from Kardar-Parisi-Zhang (KPZ) physics, dictated by the lead configurations of the scattering system and unaffected by the presence of a magnetic field. Distinguishing between these probability distributions poses a significant challenge due to their striking similarity, requiring sampling beyond the usual threshold of $\sim 10^{-6}$ accessible through independent disorder realizations. To overcome this, we implement an importance sampling scheme - a Monte Carlo approach in disorder space - that enables us to probe rare disorder configurations and sample probability distribution tails down to $10^{-30}$. This unprecedented precision allows us to unambiguously differentiate between KPZ sub-universality classes of conductance fluctuations for different lead configurations, while demonstrating the insensitivity to magnetic fields.
|
https://arxiv.org/abs/2504.17010v1
|
Poor tissue visualization and quantitative accuracy in CBCT is a major barrier in expanding the role of CBCT imaging from target localization to quantitative treatment monitoring and plan adaptations in radiation therapy sessions. To further improve image quality in CBCT, 2D antiscatter grid based scatter rejection was combined with a raw data processing pipeline and iterative image reconstruction. The culmination of these steps was referred as quantitative CBCT, qCBCT. qCBCT data processing steps include 2D antiscatter grid implementation, measurement based residual scatter, image lag, and beam hardening correction for offset detector geometry CBCT with a bow tie filter. Images were reconstructed with iterative image reconstruction to reduce image noise. To evaluate image quality, qCBCT acquisitions were performed using a variety of phantoms to investigate the effect of object size and its composition on image quality. qCBCT image quality was benchmarked against clinical CBCT and MDCT images. Addition of image lag and beam hardening correction to scatter suppression reduced HU degradation in qCBCT by 10 HU and 40 HU, respectively. When compared to gold standard MDCT, mean HU errors in qCBCT and clinical CBCT were 10 HU and 27 HU, respectively. HU inaccuracy due to change in phantom size was 22 HU and 85 HU in qCBCT and clinical CBCT images, respectively. With iterative reconstruction, contrast to noise ratio improved by a factor of 1.25 when compared to clinical CBCT protocols. Robust artifact and noise suppression in qCBCT images can reduce the image quality gap between CBCT and MDCT, improving the promise of qCBCT in fulfilling the tasks that demand high quantitative accuracy, such as CBCT based dose calculations and treatment response assessment in image guided radiation therapy.
|
https://arxiv.org/abs/2308.09095v1
|
The freshness of sensor data is critical for all types of cyber-physical systems. An established measure for quantifying data freshness is the Age-of-Information (AoI), which has been the subject of extensive research. Recently, there has been increased interest in multi-sensor systems: redundant sensors producing samples of the same physical process, sensors such as cameras producing overlapping views, or distributed sensors producing correlated samples. When the information from a particular sensor is outdated, fresh samples from other correlated sensors can be helpful. To quantify the utility of distant but correlated samples, we put forth a two-dimensional (2D) model of AoI that takes into account the sensor distance in an age-equivalent representation. Since we define 2D-AoI as equivalent to AoI, it can be readily linked to existing AoI research, especially on parallel systems. We consider physical phenomena modeled as spatio-temporal processes and derive the 2D-AoI for different Gaussian correlation kernels. For a basic exponential product kernel, we find that spatial distance causes an additive offset of the AoI, while for other kernels the effects of spatial distance are more complex and vary with time. Using our methodology, we evaluate the 2D-AoI of different spatial topologies and sensor densities.
|
https://arxiv.org/abs/2412.12789v1
|
Irregular scene text, which has complex layout in 2D space, is challenging to most previous scene text recognizers. Recently, some irregular scene text recognizers either rectify the irregular text to regular text image with approximate 1D layout or transform the 2D image feature map to 1D feature sequence. Though these methods have achieved good performance, the robustness and accuracy are still limited due to the loss of spatial information in the process of 2D to 1D transformation. Different from all of previous, we in this paper propose a framework which transforms the irregular text with 2D layout to character sequence directly via 2D attentional scheme. We utilize a relation attention module to capture the dependencies of feature maps and a parallel attention module to decode all characters in parallel, which make our method more effective and efficient. Extensive experiments on several public benchmarks as well as our collected multi-line text dataset show that our approach is effective to recognize regular and irregular scene text and outperforms previous methods both in accuracy and speed.
|
https://arxiv.org/abs/1906.05708v1
|
In two-dimensional space, we consider a system of $N$ anyons interacts via a short range attractive two-body interaction. In the stable regime, we derive the average-field Pauli functional as the mean-field limit of many-body quantum mechanics. Furthermore, we investigate the collapse phenomenon in the collapse regime where the strength of attractions tends to a critical value (defined by the cubic NLS equation) while simultaneously considering the weak field regime where the strength of the self-generated magnetic field tends to zero.
|
https://arxiv.org/abs/2409.00409v1
|
In this paper we will show that one can summarize the major two particle
reaction plane azimuthal correlations for Au + Au mid-central collisions at
$\sqrt{s_{NN}} =$ 200 GeV by defining a 2D azimuthal space which is a summary
of the event by event average.
|
http://arxiv.org/abs/1910.07597v1
|
In the era of tensions, when precision cosmology is blooming, numerous new theoretical models are emerging. However, it's crucial to pause and question the extent to which the observational data we rely on are model-dependent. In this work, we study the comoving position of the acoustic peak, a cornerstone standard ruler in cosmology. We considered BAO observational datasets from two distinct teams and calculated the product $hr_d$ with the help of each BAO data set along with SN I-a data from the Pantheon Plus sample. Our conclusion at present is that 2D and 3D BAO datasets are compatible with each other. Considering, no systematics in BAO, interpreting $\Omega_{m0}-hr_d$ plane may require physics beyond $\Lambda$CDM not just while using observational BAO data but also while observing it.
|
https://arxiv.org/abs/2406.05453v1
|
Ordinary 3D Baryon Acoustic Oscillations (BAO) data are model-dependent, requiring the assumption of a cosmological model to calculate comoving distances during data reduction. Throughout the present-day literature, the assumed model is $\Lambda$CDM. However, it has been pointed out in several recent works that this assumption can be inadequate when analyzing alternative cosmologies, potentially biasing the Hubble constant ($H_0$) low, thus contributing to the Hubble tension. To address this issue, 3D BAO data can be replaced with 2D BAO data, which is only weakly model-dependent. The impact of using 2D BAO data, in combination with alternative cosmological models beyond $\Lambda$CDM, has been explored for several phenomenological models, showing a promising reduction in the Hubble tension. In this work, we accommodate these models in the theoretically robust framework of bimetric gravity. This is a modified theory of gravity that exhibits a transition from a (possibly) negative cosmological constant in the early universe to a positive one in the late universe. By combining 2D BAO data with cosmic microwave background and type Ia supernovae data, we find that the inverse distance ladder in this theory yields a Hubble constant of $H_0 = (71.0 \pm 0.9) \, \mathrm{km/s/Mpc}$, consistent with the SH0ES local distance ladder measurement of $H_0 = (73.0 \pm 1.0) \, \mathrm{km/s/Mpc}$. Replacing 2D BAO with 3D BAO results in $H_0 = (68.6 \pm 0.5) \, \mathrm{km/s/Mpc}$ from the inverse distance ladder. Thus, the choice of BAO data significantly impacts the Hubble tension, with ordinary 3D BAO data exacerbating the tension, while 2D BAO data provides results consistent with the local distance ladder.
|
https://arxiv.org/abs/2407.04322v3
|
Basement relief gravimetry is crucial in geophysics, especially for oil exploration and mineral prospecting. It involves solving an inverse problem to infer geological model parameters from observed data. The model represents basement relief with constant-density prisms, and the data reflect gravitational anomalies from these prisms. Inverse problems are often ill-posed, meaning small data changes can lead to large solution variations. To mitigate this, regularization techniques like Tikhonov's are used to stabilize solutions. This study compares regularization methods applied to gravimetric inversion, including Smoothness Constraints, Total Variation, Discrete Cosine Transform (DCT), and Discrete Wavelet Transform (DWT) using Daubechies D4 wavelets. Optimization, particularly with Genetic Algorithms (GA), is used to find prism depths that best match observed anomalies. GA, inspired by natural selection, selects the best solutions to minimize the objective function. The results, evaluated through fit metrics and error analysis, show the effectiveness of all regularization methods and GA, with the Smoothness constraint performing best in synthetic models. For the real data model, all methods performed similarly.
|
https://arxiv.org/abs/2410.14942v1
|
We present a novel algorithm based on a Bayesian method for 2D tilted-ring analysis of disk galaxy velocity fields. Compared to the conventional algorithms based on a chi-squared minimisation procedure, this new Bayesian-based algorithm suffers less from local minima of the model parameters even with highly multi-modal posterior distributions. Moreover, the Bayesian analysis, implemented via Markov Chain Monte Carlo (MCMC) sampling, only requires broad ranges of posterior distributions of the parameters, which makes the fitting procedure fully automated. This feature will be essential when performing kinematic analysis on the large number of resolved galaxies expected to be detected in neutral hydrogen (HI) surveys with the Square Kilometre Array (SKA) and its pathfinders. The so-called '2D Bayesian Automated Tilted-ring fitter' (2DBAT) implements Bayesian fits of 2D tilted-ring models in order to derive rotation curves of galaxies. We explore 2DBAT performance on (a) artificial HI data cubes built based on representative rotation curves of intermediate-mass and massive spiral galaxies, and (b) Australia Telescope Compact Array (ATCA) HI data from the Local Volume HI Survey (LVHIS). We find that 2DBAT works best for well-resolved galaxies with intermediate inclinations (20 deg < i < 70 deg), complementing three-dimensional techniques better suited to modelling inclined galaxies.
|
https://arxiv.org/abs/1709.02049v1
|
In this paper, we investigate the beam domain statistical channel state information (CSI) estimation for the two dimensional (2D) beam based statistical channel model (BSCM) in massive MIMO systems.The problem is to estimate the beam domain channel power matrices (BDCPMs) based on multiple receive pilot signals. A receive model shows the relation between the statistical property of the receive pilot signals and the BDCPMs is derived from the 2D-BSCM. On the basis of the receive model,we formulate an optimization problem with the Kullback-Leibler (KL) divergence. By solving the optimization problem, a novel method to estimate the statistical CSI without involving instantaneous CSI is proposed. The proposed method has much lower complexity than the MMV focal underdetermined system solver (M-FOCUSS) algorithm. We further reduce the complexity of the proposed method by utilizing the circulant structures of particular matrices in the algorithm. We also showed the generality of the proposed method by introducing another application. Simulations results show that the proposed method works well and bring significant performance gain when used in channel estimation.
|
https://arxiv.org/abs/2207.04695v1
|
Many emerging reconfigurable optical systems are limited by routing complexity when producing dynamic, two-dimensional (2D) electric fields. Using a gradient-based inverse designed, static phase-mask doublet, we propose an optical system to produce 2D intensity wavefronts using a one-dimensional (1D) intensity Spatial Light Modulator (SLM). We show the capability of mapping each point in a 49 element 1D array to a distinct 7x7 2D spatial distribution. Our proposed method will significantly relax the routing complexity of 2D sub-wavelength SLMs, possibly enabling next-generation SLMs to leverage novel pixel architectures and new materials.
|
https://arxiv.org/abs/2101.04085v1
|
We report a 2D Boundary Element Method (BEM) modeling of the thermal
diffusion-controlled growth of a vapor bubble attached to a heating surface
during saturated pool boiling. The transient heat conduction problem is solved
in a liquid that surrounds a bubble with a free boundary and in a semi-infinite
solid heater. The heat generated homogeneously in the heater causes
evaporation, i. e. the bubble growth. A singularity exists at the point of the
triple (liquid-vapor-solid) contact. At high system pressure the bubble is
assumed to grow slowly, its shape being defined by the surface tension and the
vapor recoil force, a force coming from the liquid evaporating into the bubble.
It is shown that at some typical time the dry spot under the bubble begins to
grow rapidly under the action of the vapor recoil. Such a bubble can eventually
spread into a vapor film that can separate the liquid from the heater, thus
triggering the boiling crisis (Critical Heat Flux phenomenon).
|
http://arxiv.org/abs/1601.07196v1
|
Abnormal behavior detection, action recognition, fight and violence detection in videos is an area that has attracted a lot of interest in recent years. In this work, we propose an architecture that combines a Bidirectional Gated Recurrent Unit (BiGRU) and a 2D Convolutional Neural Network (CNN) to detect violence in video sequences. A CNN is used to extract spatial characteristics from each frame, while the BiGRU extracts temporal and local motion characteristics using CNN extracted features from multiple frames. The proposed end-to-end deep learning network is tested in three public datasets with varying scene complexities. The proposed network achieves accuracies up to 98%. The obtained results are promising and show the performance of the proposed end-to-end approach.
|
https://arxiv.org/abs/2409.07588v1
|
We report two surprising results on $\alpha'$ corrections in string theory restricted to massless fields. First, for critical dimension Bianchi type I cosmologies with $q$ scale factors only $q-1$ of them have non-trivial $\alpha'$ corrections. In particular, for FRW backgrounds all $\alpha'$ corrections are trivial. Second, in non-critical dimensions, all terms in the spacetime action other than the cosmological term are field redefinition equivalent to terms with arbitrarily many derivatives, with the latter generally of the same order. Assuming an $\alpha'$ expansion with coefficients that fall off sufficiently fast, we consider field redefinitions consistent with this fall-off and classify the higher derivative terms for two-dimensional string theory with one timelike isometry. This most general duality-invariant theory permits black-hole solutions, and we provide perturbative and non-perturbative tools to explore them.
|
https://arxiv.org/abs/2304.06763v2
|
The allotropes of a new layered material, phosphorus carbide (PC), have been
predicted recently and a few of these predicted structures have already been
successfully fabricated. Herein, by using first-principles calculations we
investigated the effects of rippling a PC monolayer, one of the most stable
modifications of layered PC, under large compressive strains. Similar to
phosphorene, layered PC was found to have the extraordinary ability to bend and
form ripples with large curvatures under a sufficiently large strain applied
along its armchair direction. The band gap size, workfunction, and Young's
modulus of rippled PC monolayer are predicted to be highly tunable by strain
engineering. Moreover, a direct-indirect band gap transition is observed under
the compressive strains in a range from 6 to 11%. Another important feature of
PC monolayer rippled along the armchair direction is the possibility of its
rolling to a PC nanotube (PCNT) under extreme compressive strains. These tubes
of different sizes exhibit high thermal stability, possess a comparably high
Young's modulus, and a well tunable band gap which can vary from 0 to 0.95 eV.
In addition, for both structures, rippled PC and PCNTs, we have explained the
changes in their properties under compressive strain in terms of the
modification of their structural parameters.
|
http://arxiv.org/abs/2002.08093v1
|
We present a new geminal product wave function ansatz where the geminals are not constrained to be strongly orthogonal nor to be of seniority zero. Instead, we introduce weaker orthogonality constraints between geminals which significantly lower the computational effort, without sacrificing the indistinguishability of the electrons. That is to say, the electron pairs corresponding to the geminals are not fully distinguishable, and their product has still to be antisymmetrized according to the Pauli principle to form a \textit{bona fide} electronic wave function.Our geometrical constraints translate into simple equations involving the traces of products of our geminal matrices. In the simplest non-trivial model, a set of solutions is given by block-diagonal matrices where each block is of size 2x2 and consists of either a Pauli matrix or a normalized diagonal matrix, multiplied by a complex parameter to be optimized. With this simplified ansatz for geminals, the number of terms in the calculation of the matrix elements of quantum observables is considerably reduced. A proof of principle is reported and confirms that the ansatz is more accurate than strongly orthogonal geminal products while remaining computationally affordable.
|
https://arxiv.org/abs/2209.00834v4
|
Multiple dissipative self-assembly protocols designed to create novel structures or to reduce kinetic traps have recently emerged. Specifically, temporal oscillations of particle interactions have been shown effective at both aims, but investigations thus far have focused on systems of simple colloids or their binary mixtures. In this work, we expand our understanding of the effect of temporally oscillating interactions to a two-dimensional coarse-grained viral capsid-like model that undergoes a self-limited assembly. This model includes multiple intrinsic relaxation times due to the internal structure of the capsid subunits and, under certain interaction regimes, proceeds via a two-step nucleation mechanism. We find that oscillations much faster than the local intrinsic relaxation times can be described via a time averaged inter-particle potential across a wide range of interaction strengths, while oscillations much slower than these relaxation times result in structures that adapt to the attraction strength of the current half-cycle. Interestingly, oscillation periods similar to these relaxation times shift the interaction window over which orderly assembly occurs by enabling error correction during the half-cycles with weaker attractions. Our results provide fundamental insights to non-equilibrium self-assembly on temporally variant energy landscapes.
|
https://arxiv.org/abs/2404.11765v2
|
For many automated driving functions, a highly accurate perception of the vehicle environment is a crucial prerequisite. Modern high-resolution radar sensors generate multiple radar targets per object, which makes these sensors particularly suitable for the 2D object detection task. This work presents an approach to detect 2D objects solely depending on sparse radar data using PointNets. In literature, only methods are presented so far which perform either object classification or bounding box estimation for objects. In contrast, this method facilitates a classification together with a bounding box estimation of objects using a single radar sensor. To this end, PointNets are adjusted for radar data performing 2D object classification with segmentation, and 2D bounding box regression in order to estimate an amodal 2D bounding box. The algorithm is evaluated using an automatically created dataset which consist of various realistic driving maneuvers. The results show the great potential of object detection in high-resolution radar data using PointNets.
|
https://arxiv.org/abs/1904.08414v3
|
The Cardy-Rabinovici model is a toy model of the lattice $U(1)$ gauge theories to study various oblique confinement states associated with the nonzero $\theta$ angles. We reformulate the $2$d version of this model using the modified Villain lattice formalism, and we establish the exact $\theta$ periodicity for the Witten effect and the strong-weak duality at the finite lattice spacings. We then study the phase structure of this model based on the duality, symmetry and anomaly, and the perturbative renormalization group.
|
https://arxiv.org/abs/2505.19412v1
|
We study a 2D cellular automaton (CA) model for the evolution of coronal loop
plasmas. The model is based on the idea that coronal loops are made of
elementary magnetic strands that are tangled and stressed by the displacement
of their footpoints by photospheric motions. The magnetic stress accumulated
between neighbor strands is released in sudden reconnection events or
nanoflares that heat the plasma. We combine the CA model with the Enthalpy
Based Thermal Evolution of Loops (EBTEL) model to compute the response of the
plasma to the heating events. Using the known response of the XRT telescope on
board Hinode we also obtain synthetic data. The model obeys easy to understand
scaling laws relating the output (nanoflare energy, temperature, density,
intensity) to the input parameters (field strength, strand length, critical
misalignment angle). The nanoflares have a power-law distribution with a
universal slope of -2.5, independent of the input parameters. The repetition
frequency of nanoflares, expressed in terms of the plasma cooling time,
increases with strand length. We discuss the implications of our results for
the problem of heating and evolution of active region coronal plasmas.
|
http://arxiv.org/abs/1607.03883v1
|
We consider the late time behavior of the analytically continued partition
function $Z(\beta + it) Z(\beta - it)$ in holographic $2d$ CFTs. This is a
probe of information loss in such theories and in their holographic duals. We
show that each Virasoro character decays in time, and so information is not
restored at the level of individual characters. We identify a universal
decaying contribution at late times, and conjecture that it describes the
behavior of generic chaotic $2d$ CFTs out to times that are exponentially large
in the central charge. It was recently suggested that at sufficiently late
times one expects a crossover to random matrix behavior. We estimate an upper
bound on the crossover time, which suggests that the decay is followed by a
parametrically long period of late time growth. Finally, we discuss integrable
theories and show how information is restored at late times by a series of
characters. This hints at a possible bulk mechanism, where information is
restored by an infinite sum over non-perturbative saddles.
|
http://arxiv.org/abs/1611.04592v1
|
In this note we study two-dimensional CFTs at large global charge. Since the large-charge sector decouples from the dynamics, it does not control the dynamics and an EFT construction that works in higher-dimensional theories fails. It is however possible to use large charge in a double-scaling limit when another controlling parameter is present. We find some general features of the spectrum of models that admit an NLSM description in a WKB approximation and use the large-charge sector of the solvable $SU(2)_k$ WZW model to argue the regimes of applicability of both the large-Q expansion and the double-scaling limit.
|
https://arxiv.org/abs/2112.03286v2
|
According to observations and numerical simulations, the Milky Way could exhibit several spiral arm modes with multiple pattern speeds, wherein the slower patterns are located at larger Galactocentric distances. Our aim is to quantify the effects of the spiral arms on the azimuthal variations of the chemical abundances for oxygen, iron and for the first time for neutron-capture elements (europium and barium) in the Galactic disc. We assume a model based on multiple spiral arm modes with different pattern speeds. The resulting model represents an updated version of previous 2D chemical evolution models. We apply new analytical prescriptions for the spiral arms in a 2D Galactic disc chemical evolution model, exploring the possibility that the spiral structure is formed by the overlap of chunks with different pattern speeds and spatial extent. The predicted azimuthal variations in abundance gradients are dependent on the considered chemical element. Elements synthesised on short time scales (i.e., oxygen and europium in this study) exhibit larger abundance fluctuations. In fact, for progenitors with short lifetimes, the chemical elements restored into the ISM perfectly trace the star formation perturbed by the passage of the spiral arms. The map of the star formation rate predicted by our chemical evolution model with multiple patterns of spiral arms presents arcs and arms compatible with those revealed by multiple tracers (young upper main sequence stars, Cepheids, and distribution of stars with low radial actions). Finally, our model predictions are in good agreement with the azimuthal variations that emerged from the analysis of Gaia DR3 GSP-Spec [M/H] abundance ratios, if at most recent times the pattern speeds match the Galactic rotational curve at all radii.
|
https://arxiv.org/abs/2310.11504v1
|
Galactic disc chemical evolution models generally ignore azimuthal surface
density variation that can introduce chemical abundance azimuthal gradients.
Recent observations, however, have revealed chemical abundance changes with
azimuth in the gas and stellar components of both the Milky Way and external
galaxies. To quantify the effects of spiral arm density fluctuations on the
azimuthal variations of the oxygen and iron abundances in disc galaxies. We
develop a new 2D galactic disc chemical evolution model, capable of following
not just radial but also azimuthal inhomogeneities. The density fluctuations
resulting from a Milky Way-like N-body disc formation simulation produce
azimuthal variations in the oxygen abundance gradients of the order of 0.1 dex.
Moreover, in agreement with the most recent observations in external galaxies,
the azimuthal variations are more evident in the outer galactic regions. Using
a simple analytical model, we show that the largest fluctuations with azimuth
result near the spiral structure corotation resonance, where the relative speed
between spiral and gaseous disc is the slowest. In conclusion we provided a new
2D chemical evolution model capable of following azimuthal density variations.
Density fluctuations extracted from a Milky Way-like dynamical model lead to a
scatter in the azimuthal variations of the oxygen abundance gradient in
agreement with observations in external galaxies. We interpret the presence of
azimuthal scatter at all radii by the presence of multiple spiral modes moving
at different pattern speeds, as found in both observations and numerical
simulations.
|
http://arxiv.org/abs/1811.11196v3
|
Despite being invented in 1951 by R. Kikuchi, the 2-D Cluster Variation Method (CVM), has not yet received attention. Nevertheless, this method can usefully characterize 2-D topographies using just two parameters; the activation enthalpy and the interaction enthalpy. This Technical Report presents 2-D CVM details, including the dependence of the various configuration variables on the enthalpy parameters, as well as illustrations of various topographies (ranging from scale-free-like to rich club-like) that result from different parameter selection. The complete derivation for the analytic solution, originally presented simply as a result in Kikuchi and Brush (1967) is given here, along with careful comparison of the analytically-predicted configuration variables versus those obtained when performing computational free energy minimization on a 2-D grid. The 2-D CVM can potentially function as a secondary free energy minimization within the hidden layer of a neural network, providing a basis for extending node activations over time and allowing temporal correlation of patterns.
|
https://arxiv.org/abs/1909.09366v1
|
The coherence factor (CF) is defined as the ratio of coherent power to
incoherent power received by the radar aperture. The incoherent power is
computed by the multi-antenna receiver based on only the spatial variable. In
this respect, it is a one-dimensional (1-D) CF, and thereby the image sidelobes
in down-range cannot be effectively suppressed. We propose a two-dimensional
(2-D) CF by supplementing the 1-D CF by an incoherent sum dealing with the
frequency dimension. In essence, we employ both spatial diversity and frequency
diversity which, respectively, enhance imaging quality in cross range and
range. Simulations and experimental results are provided to demonstrate the
performance advantages of the proposed approach.
|
http://arxiv.org/abs/1903.10119v1
|
The compass model on a square lattice provides a natural template for
building subsystem stabilizer codes. The surface code and the Bacon-Shor code
represent two extremes of possible codes depending on how many gauge qubits are
fixed. We explore threshold behavior in this broad class of local codes by
trading locality for asymmetry and gauge degrees of freedom for stabilizer
syndrome information. We analyze these codes with asymmetric and spatially
inhomogeneous Pauli noise in the code capacity and phenomenological models. In
these idealized settings, we observe considerably higher thresholds against
asymmetric noise. At the circuit level, these codes inherit the bare-ancilla
fault-tolerance of the Bacon-Shor code.
|
http://arxiv.org/abs/1809.01193v2
|
The electron-hole liquid, which features a macroscopic population of
correlated electrons and holes, may offer a path to room temperature
semiconductor devices that harness collective electronic phenomena. We report
on the gas-to-liquid phase transition of electrons and holes in ultrathin
molybdenum ditelluride photocells revealed through multi-parameter dynamic
photoresponse microscopy (MPDPM). By combining rich visualization with
comprehensive analysis of very large data sets acquired through MPDPM, we find
that ultrafast laser excitation at a graphene-MoTe$_2$-graphene interface leads
to the abrupt formation of ring-like spatial patterns in the photocurrent
response as a function of increasing optical power at T = 297 K. These
patterns, together with extreme sublinear power dependence and picosecond-scale
photocurrent dynamics, provide strong evidence for the formation of a
two-dimensional electron-hole condensate.
|
http://arxiv.org/abs/1711.06917v1
|
We study 2D Navier-Stokes equations with a constraint on $L^2$ energy of the
solution. We prove the existence and uniqueness of a global solution for the
constrained Navier-Stokes equation on $\R^2$ and $\T$, by a fixed point
argument. We also show that the solution of constrained Navier-Stokes converges
to the solution of Euler equation as viscosity $\nu$ vanishes.
|
http://arxiv.org/abs/1606.08360v1
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.