text
stringlengths
27
153
label
stringlengths
27
153
id
int64
0
40k
and network that synCTs allows sampling multiple This their representation. share segmentations of
This allows sampling of multiple segmentations and synCTs that share their network representation.
36,500
for networks multi-focus fusion. Convolutional used been image neural have recently
Convolutional neural networks have recently been used for multi-focus image fusion.
36,501
vector popular recognition spaces a strategy pattern Embedding very data methods. into of is
Embedding data into vector spaces is a very popular strategy of pattern recognition methods.
36,502
ambiguous. distances metrics When embeddings are between quantized, performance become
When distances between embeddings are quantized, performance metrics become ambiguous.
36,503
systems. a it measurable state-of-the-art that have demonstrate in can empirical data in effect We
We demonstrate that it can have a measurable effect in empirical data in state-of-the-art systems.
36,504
this model, of Point-of-care at roles. systems, two important the play heart
Point-of-care systems, at the heart of this model, play two important roles.
36,505
will the be of assessing progression such used they Second, therapies. for
Second, they will be used for assessing the progression of such therapies.
36,506
deconvolution local detection. present alternate We to as an approach maxima
We present deconvolution as an alternate approach to local maxima detection.
36,507
a are filter truth artifical points generate convolved The ground with to labels. mapping
The ground truth points are convolved with a mapping filter to generate artifical labels.
36,508
CNN cell generate is detection. of then deconvolved as the to Output trained points
Output of the trained CNN is then deconvolved to generate points as cell detection.
36,509
directions for these rewards. we leveraging further discuss Finally, intrinsic between the future interplay two
Finally, we discuss future directions for further leveraging the interplay between these two intrinsic rewards.
36,510
are methods surgical for achieve To workflow analysis crucial. this,
To achieve this, methods for surgical workflow analysis are crucial.
36,511
video-based the performance best convolutional for Currently, networks tasks. workflow analysis neural provide
Currently, convolutional neural networks provide the best performance for video-based workflow analysis tasks.
36,512
data annotated For of such networks, training necessary. large amounts are
For training such networks, large amounts of annotated data are necessary.
36,513
always a feasible. and time-consuming, not costly, of amount is collecting sufficient often However, data
However, collecting a sufficient amount of data is often costly, time-consuming, and not always feasible.
36,514
one system simulating conditioned consists particle path. particle step of a on Each Gibbs
Each Gibbs step consists of simulating a particle system conditioned on one particle path.
36,515
Monte to method create the system. conditional a Sequential (cSMC) relies It particle on Carlo
It relies on a conditional Sequential Monte Carlo (cSMC) method to create the particle system.
36,516
can a This get used uniform result. ergodicity to be
This can be used to get a uniform ergodicity result.
36,517
continuum. modelling cell either typically macro-scale, are to two There discrete micro-scale, or approaches migration:
There are typically two approaches to modelling cell migration: either micro-scale, discrete or macro-scale, continuum.
36,518
mathematical have Deterministic advantage to generally analysis. they more models are the amenable that
Deterministic models have the advantage that they are generally more amenable to mathematical analysis.
36,519
phenomenological predictive their models power. often specified which Deterministic on are basis, reduces a
Deterministic models are often specified on a phenomenological basis, which reduces their predictive power.
36,520
continuum ABMs over deterministic Stochastic advantages have their counterparts.
Stochastic ABMs have advantages over their deterministic continuum counterparts.
36,521
actions. This the procedures allows consisting multiple of of derivation
This allows the derivation of procedures consisting of multiple actions.
36,522
Tomography (PET) Positron as functional from perfusion and metabolism. such Images Emission deliver data
Images from Positron Emission Tomography (PET) deliver functional data such as perfusion and metabolism.
36,523
Imaging structures. hand, Resonance (MRI) provide from Magnetic the information other On describing anatomical images
On the other hand, images from Magnetic Resonance Imaging (MRI) provide information describing anatomical structures.
36,524
Fusing is from two the helpful in complementary the information oncology. modalities
Fusing the complementary information from the two modalities is helpful in oncology.
36,525
hierarchical A two-stage approach proposed. is registration
A two-stage hierarchical registration approach is proposed.
36,526
global a is applied. registration affine First,
First, a global affine registration is applied.
36,527
optimization. as function Mutual used metric the for information is
Mutual information is used as metric function for the optimization.
36,528
stages. in is both A used approach multi-resolution
A multi-resolution approach is used in both stages.
36,529
in world. our is repetition Visual ubiquitous
Visual repetition is ubiquitous in our world.
36,530
continuities temporal field's three addition, dynamics. from the motion temporal emerge In
In addition, three temporal motion continuities emerge from the field's temporal dynamics.
36,531
repetitive filters dynamics. the estimate We wavelet motion with to maps temporally convolve
We temporally convolve the motion maps with wavelet filters to estimate repetitive dynamics.
36,532
rotor resulted targeting in termination. Ablation sites arrhythmia has
Ablation targeting rotor sites has resulted in arrhythmia termination.
36,533
to However rotors leading are not abundance at clear. mechanisms locations the of these
However the mechanisms leading to abundance of rotors at these locations are not clear.
36,534
with change from This monomorphic ventricular process coalesces tachycardia. to polymorphic a
This process coalesces with a change from polymorphic to monomorphic ventricular tachycardia.
36,535
estimation. proposed approach on for available datasets Our pose is evaluated publicly several
Our proposed approach is evaluated on several publicly available datasets for pose estimation.
36,536
classification model and semi-supervised on our settings. unsupervised in localization We and also evaluated
We also evaluated our model on classification and localization in unsupervised and semi-supervised settings.
36,537
retrieval because tasks techniques hashing retrieval of their gained large-scale speed. Recently, have importance in
Recently, hashing techniques have gained importance in large-scale retrieval tasks because of their retrieval speed.
36,538
data cross-view existing paired. frameworks assume the Most are well of that
Most of the existing cross-view frameworks assume that data are well paired.
36,539
situation in applications. is However, fully-paired not universal real multiview the
However, the fully-paired multiview situation is not universal in real applications.
36,540
detection under presence missing robustness annotations. of object the study of the We
We study the robustness of object detection under the presence of missing annotations.
36,541
detailed for this We explanation provide result. a
We provide a detailed explanation for this result.
36,542
a re-weights gradients instances. of function the overlap positive RoIs Sampling with of Soft as
Soft Sampling re-weights the gradients of RoIs as a function of overlap with positive instances.
36,543
HexaLab by simply opening used can be www.hexalab.net.
HexaLab can be used by simply opening www.hexalab.net.
36,544
tool both visualization targets scholars. users and Our
Our visualization tool targets both users and scholars.
36,545
inspection support of variety end, we this visualization and a To wide volume tools.
To this end, we support a wide variety of visualization and volume inspection tools.
36,546
People sensing issues one hottest applications. counting of in is the
People counting is one of the hottest issues in sensing applications.
36,547
occlusion shading network. and calculation encoded in automatically the Spatial are
Spatial occlusion and shading calculation are automatically encoded in the network.
36,548
feature used perturbation this chain assess be may robustness. to Thus,
Thus, this perturbation chain may be used to assess feature robustness.
36,549
heart the lies tracking. of Model at update object
Model update lies at the heart of object tracking.
36,550
training CNNs However, requires annotated image data.
However, training CNNs requires annotated image data.
36,551
emphysema of CT We scans. consider in chest the assessing extent task
We consider the task of assessing emphysema extent in chest CT scans.
36,552
how our CAPTCHAs. these are system towards work provided well defeating showing Results
Results are provided showing how our system work well towards defeating these CAPTCHAs.
36,553
continuously systems physical interact complex the robotic dynamical systems with The in world.
The robotic systems continuously interact with complex dynamical systems in the physical world.
36,554
classification. algorithms on discovery poorly Unfortunately, fully-automated score this state-of-the-art even the
Unfortunately, even the state-of-the-art fully-automated discovery algorithms score poorly on this classification.
36,555
at aims work addressing This points. two those
This work aims at addressing those two points.
36,556
the first identifies approach of GBM. two-stage presence A
A two-stage approach first identifies the presence of GBM.
36,557
slice. each "abnormal" in localization MR is followed GBM a This by
This is followed by a GBM localization in each "abnormal" MR slice.
36,558
viz. As two system, of architectures CADe part the CNN
As part of the CADe system, two CNN architectures viz.
36,559
Detection (C-CNN) and are CNN Classification employed. CNN (D-CNN)
Classification CNN (C-CNN) and Detection CNN (D-CNN) are employed.
36,560
of surgery assessment surgical efficient for Recognition training. and is surgical crucial gesture skill
Recognition of surgical gesture is crucial for surgical skill assessment and efficient surgery training.
36,561
usually current segment-level from Most edit the therefore over-segmentation of scores. low approaches suffer and
Most of the current approaches usually suffer from over-segmentation and therefore low segment-level edit scores.
36,562
deep from a learning features with trained intelligent using hierarchical is agent reinforcement model. An
An intelligent agent is trained using reinforcement learning with hierarchical features from a deep model.
36,563
consistency and mechanism Temporal action into design over-segmentation reduce errors. reward our integrated is to
Temporal consistency is integrated into our action design and reward mechanism to reduce over-segmentation errors.
36,564
approaches of Unfortunately, existing problem. from the same majority suffers the
Unfortunately, the majority of existing approaches suffers from the same problem.
36,565
CNNs I present use In i.e. this of another paper
In this paper I present another use of CNNs i.e.
36,566
and detectors on light perception. integral cars Traffic for autonomous scene sign are road
Traffic light and sign detectors on autonomous cars are integral for road scene perception.
36,567
region boxes, proposed method of or bounding notion has sliding no windows. proposals, The
The proposed method has no notion of bounding boxes, region proposals, or sliding windows.
36,568
object detectors fine-tuned and We methods state-of-the-art tracking. outperform for pupil generic
We outperform state-of-the-art generic object detectors and methods fine-tuned for pupil tracking.
36,569
This paper EH-powered some theoretical CPS architecture the problems. discusses and
This paper discusses the EH-powered CPS architecture and some theoretical problems.
36,570
partially EH EH-powered and fully addressed. (e.g., CPSs Different for models EH-powered) are
Different EH models for CPSs (e.g., fully EH-powered and partially EH-powered) are addressed.
36,571
directions discussed research in the Future are end. briefly
Future research directions are briefly discussed in the end.
36,572
not However, representations. do most deep networks feature such enforce
However, most deep networks do not enforce such feature representations.
36,573
important it not is per In the is candidate. performance context, this what
In this context, it is not important what is the performance per candidate.
36,574
Instead objective. group per the the ultimate performance is
Instead the performance per group is the ultimate objective.
36,575
millions algorithm of was handle hundreds of The to designed examples. efficiently
The algorithm was designed to efficiently handle hundreds of millions of examples.
36,576
the being automatic capture is virtual environment is screen Concurrently, executed while explored.
Concurrently, automatic screen capture is executed while the virtual environment is being explored.
36,577
We self-supervised using spatio-temporal between frames for signals action present recognition. a video approach
We present a self-supervised approach using spatio-temporal signals between video frames for action recognition.
36,578
tangle architecture to A temporal is and spatial representation two-stream learning. leveraged
A two-stream architecture is leveraged to tangle spatial and temporal representation learning.
36,579
is as verification Our and a task formulated tasks. both sequence spatio-temporal alignment
Our task is formulated as both a sequence verification and spatio-temporal alignment tasks.
36,580
The action weights self-supervised validated is on the effectiveness pre-trained task. recognition
The self-supervised pre-trained weights effectiveness is validated on the action recognition task.
36,581
are to and still validity boost required. investigations Further generalize performance
Further investigations to boost performance and generalize validity are still required.
36,582
MRIs of plausibility this gate we In the for paper, using diagnosing PD. automatically investi-
In this paper, we investi- gate the plausibility of using MRIs for automatically diagnosing PD.
36,583
the FreeSurfer is second and library the used steps. for first The
The FreeSurfer library is used for the first and the second steps.
36,584
evaluate method. set used proposed to Markers data the Progression Parkinsons is The (PPMI) Initiative
The Parkinsons Progression Markers Initiative (PPMI) data set is used to evaluate the proposed method.
36,585
system be The proposed the in of PD. prove promising diagnosis assisting to
The proposed system prove to be promising in assisting the diagnosis of PD.
36,586
in prerequisites appropriate provided the mathematical are The appendices.
The appropriate mathematical prerequisites are provided in the appendices.
36,587
as used (SMO). Spectral Sequential features Optimization input features have for based Minimal been
Spectral based features have been used as input features for Sequential Minimal Optimization (SMO).
36,588
segmentation tasks. Random image medical forests (RF) successfully applied many have to been
Random forests (RF) have been successfully applied to many medical image segmentation tasks.
36,589
pixels. contextual pixel-wise However, relationships the outputs ignores between classifier label RF of individual
However, the pixel-wise RF classifier ignores contextual relationships between label outputs of individual pixels.
36,590
transformation explore using as ITN. effect representations different regression of outputs We the of
We explore the effect of using different transformation representations as regression outputs of ITN.
36,591
complete to are manipulation Also, subtle most involved there actions. object of movements
Also, there are subtle movements involved to complete most of object manipulation actions.
36,592
determine to i.e. graph the p-Lapalcian fitting is difficult it However,
However, it is difficult to determine the fitting graph p-Lapalcian i.e.
36,593
parameter performance the is graph factor the critical a which p-Laplacian. for of
the parameter which is a critical factor for the performance of graph p-Laplacian.
36,594
on the conduct then, graph. learning And framework semi-supervised fused we
And then, we conduct semi-supervised learning framework on the fused graph.
36,595
the data and set Extensive efficiency UC-Merced proposed demonstrate the experiments method. on of effectiveness
Extensive experiments on UC-Merced data set demonstrate the effectiveness and efficiency of the proposed method.
36,596
available new, on In text/video a to we large-scale datasets, dataset. publicly introduce testing addition
In addition to testing on publicly available datasets, we introduce a new, large-scale text/video dataset.
36,597
(FPS) First-Person in the game. We the of Shooter technique test context a
We test the technique in the context of a First-Person Shooter (FPS) game.
36,598
closed is scene images, once, to allowing such geometries. recover at With seen all
With such images, all scene is seen at once, allowing to recover closed geometries.
36,599