text
stringlengths
27
153
label
stringlengths
27
153
id
int64
0
40k
infer is lineage from data the to alternative phenotype estimation. latent-variable An approach the by
An alternative approach is to infer the phenotype from the lineage data by latent-variable estimation.
35,900
statistical work, how this we estimations. distorts In the survivor clarify bias
In this work, we clarify how the survivor bias distorts statistical estimations.
35,901
TNM the determinant and prognosis treatment. Predicting is stage of cancer breast major
Predicting TNM stage is the major determinant of breast cancer prognosis and treatment.
35,902
detecting Pathologic commonly histological in (pN-stage) metastasis by slides. N-stage pathologists performed is
Pathologic N-stage (pN-stage) is commonly performed by pathologists detecting metastasis in histological slides.
35,903
by is classifier. combining metastasis node CNN lymph patch-level pN-stage based and detector slide-level predicted
pN-stage is predicted by combining patch-level CNN based metastasis detector and slide-level lymph node classifier.
35,904
denoising LF parallax propose on anisotropic a framework based We novel (APA). analysis
We propose a novel LF denoising framework based on anisotropic parallax analysis (APA).
35,905
to reinforcement the video learn Humans play algorithms. significantly (RL) state-of-the-art than learning faster games
Humans learn to play video games significantly faster than the state-of-the-art reinforcement learning (RL) algorithms.
35,906
efficiency. we Inspired model-based leveraging sample investigate in by for two issues this, RL
Inspired by this, we investigate two issues in leveraging model-based RL for sample efficiency.
35,907
suffers from output This ineffiency can discontinuities at result computational and boundaries. approach in
This approach suffers from computational ineffiency and can result in discontinuities at output boundaries.
35,908
substantially mitigates not does but avoid This limitations. stitching altogether, its
This does not avoid stitching altogether, but substantially mitigates its limitations.
35,909
(overall approach competition. labeling our in entry performance) to the contributed This wining building INRIA
This approach contributed to our wining entry (overall performance) in the INRIA building labeling competition.
35,910
are exploring visual of current on on Most focusing features. recognition researches emotion facial
Most of current researches on visual emotion recognition are focusing on exploring facial features.
35,911
detecting our We test method the pelvic frontal task on hip of fractures from x-rays.
We test our method on the task of detecting hip fractures from frontal pelvic x-rays.
35,912
of compression a BPG artifact removal method propose images. We compressed learned for post-processing by
We propose a method for learned compression artifact removal by post-processing of BPG compressed images.
35,913
networks trained three different We of sizes.
We trained three networks of different sizes.
35,914
We encoded images input different using with QP values. BPG
We encoded input images using BPG with different QP values.
35,915
combination is an the integer posed best as problem. programming The of selection
The selection of the best combination is posed as an integer programming problem.
35,916
imaging in diffraction-photons. problem This paper addresses the of presence the of
This paper addresses the problem of imaging in the presence of diffraction-photons.
35,917
experiment conducted demonstrate of a Additionally, designed proof-of-concept the to the scheme. is viability
Additionally, a proof-of-concept experiment is conducted to demonstrate the viability of the designed scheme.
35,918
has dermoscopic analysis image rapid growth performance. in Automated diagnostic witnessed
Automated dermoscopic image analysis has witnessed rapid growth in diagnostic performance.
35,919
evidence faces decisions. part, in provided adoption support is because resistance, to no Yet
Yet adoption faces resistance, in part, because no evidence is provided to support decisions.
35,920
work, an is this for classification approach In evidence-based presented.
In this work, an approach for evidence-based classification is presented.
35,921
optimal parameters size, etc.) rate, (dilution choosing volume operational Moreover,
Moreover, choosing optimal operational parameters (dilution rate, volume size, etc.)
35,922
with different learning domains. are generalize well concerned models \emph{unseen} to that We
We are concerned with learning models that generalize well to different \emph{unseen} domains.
35,923
novel (GANs). unsupervised Networks approach Generative a introduces Adversarial using co-localization paper for This object
This paper introduces a novel approach for unsupervised object co-localization using Generative Adversarial Networks (GANs).
35,924
consider problem the of summarization. video We
We consider the problem of video summarization.
35,925
such it very However, labeled to and difficult training expensive is examples. create
However, it is very expensive and difficult to create such labeled training examples.
35,926
is this collect. type argue much that to We easier of data
We argue that this type of data is much easier to collect.
35,927
integration. brain and successfully such Our segregation functional important ERGMs reproduced as pooled functional properties
Our pooled ERGMs successfully reproduced important brain properties such as functional segregation and functional integration.
35,928
improve and robustness a with Our larger sample promising are size. in would results
Our results are promising and would improve in robustness with a larger sample size.
35,929
with confounding or variables, hidden (models with insufficient or etc.) latent Causally structures
Causally insufficient structures (models with latent or hidden variables, or with confounding etc.)
35,930
developed the data. belief been from Algorithms network to acquire structure therefore have
Algorithms have been therefore developed to acquire the belief network structure from data.
35,931
already Regrettably, erroneous them are some (e.g. be to of known
Regrettably, some of them are known already to be erroneous (e.g.
35,932
this described the for in paper. case
for the case described in this paper.
35,933
this proposed. remedy is of A failure
A remedy of this failure is proposed.
35,934
validate method through We computer simulations. the new
We validate the new method through computer simulations.
35,935
the considerably healthy Additionally, alters statistics network find we controls. of aripiprazole that
Additionally, we find that aripiprazole considerably alters the network statistics of healthy controls.
35,936
establish aripiprazole their test performance. of adversely we also cognitive a that Using ability, affects
Using a test of cognitive ability, we establish that aripiprazole also adversely affects their performance.
35,937
evidence that network brain This measurable result differences. architecture macroscopic in to provides changes behavioural
This provides evidence that changes to macroscopic brain network architecture result in measurable behavioural differences.
35,938
is This different assessed medications time in have been the first way. this
This is the first time different medications have been assessed in this way.
35,939
averaging multiple the acquiring techniques to improve rely and Current signal acquisitions SNR. DT-CMR on
Current DT-CMR techniques rely on acquiring and averaging multiple signal acquisitions to improve the SNR.
35,940
It interaction, and has visually subtitling. in the video human-robot applications impaired helping
It has applications in human-robot interaction, helping the visually impaired and video subtitling.
35,941
nor visual neither diversity of linguistic datasets complexity adequate Existing structures. contain
Existing datasets neither contain adequate visual diversity nor complexity of linguistic structures.
35,942
introduced. briefly representation which eight six Omega of and networks, languages are classes includes neural
Omega includes eight representation languages and six classes of neural networks, which are briefly introduced.
35,943
ontologies systems. still ontology challenges serious state-of-the-art alignment Large pose to
Large ontologies still pose serious challenges to state-of-the-art ontology alignment systems.
35,944
Evaluation evaluation have of datasets Alignment a conducted comprehensive We the Ontology using the Initiative.
We have conducted a comprehensive evaluation using the datasets of the Ontology Alignment Evaluation Initiative.
35,945
propose tracking, communications. For new we model a mmWave vehicular beam practical for
For beam tracking, we propose a new practical model for mmWave vehicular communications.
35,946
accurate more show produces that optic we traditional than algorithms. SOFAS results flow Finally,
Finally, we show that SOFAS produces more accurate results than traditional optic flow algorithms.
35,947
is the a efficient simple lower EE obtained. Furthermore, bound yet of
Furthermore, a simple yet efficient lower bound of the EE is obtained.
35,948
consists usually weeds in modern agriculture, In the over field. spraying all control herbicides agricultural
In modern agriculture, usually weeds control consists in spraying herbicides all over the agricultural field.
35,949
This herbicide farmers and of environmental practice and pollution. waste cost significant for involves
This practice involves significant waste and cost of herbicide for farmers and environmental pollution.
35,950
Learning impressive problem. different complex Recently shown approach in Deep classification has results
Recently Deep Learning approach has shown impressive results in different complex classification problem.
35,951
method three main consists phases. in proposed The
The proposed method consists in three main phases.
35,952
automatically identify interline to the using detect lines them and crop we First weeds. the
First we automatically detect the crop lines and using them to identify the interline weeds.
35,953
used In second interline phase, to constitute dataset. training are weeds the the
In the second phase, interline weeds are used to constitute the training dataset.
35,954
comparable supervised obtained traditional results training are the labeling. data The to
The results obtained are comparable to the traditional supervised training data labeling.
35,955
when produce divergence characterizes error. view keypoint Epipolar zero distributions reprojection two
Epipolar divergence characterizes when two view keypoint distributions produce zero reprojection error.
35,956
undergone recognition has few adaptation visual progress the in for Domain past great years.
Domain adaptation for visual recognition has undergone great progress in the past few years.
35,957
a nonmonotonic Programming based well-known on Answer is approach Set programs. solving problem (ASP) logic
Answer Set Programming (ASP) is a well-known problem solving approach based on nonmonotonic logic programs.
35,958
this answer inconsistent consider sets. without HEX-programs, In programs work and we ASP- i.e.,
In this work we consider inconsistent ASP- and HEX-programs, i.e., programs without answer sets.
35,959
two theoretical motivated by we This applications, concrete work which is also present.
This theoretical work is motivated by two concrete applications, which we also present.
35,960
for a algorithm a single is Typically, used. ground-truth single
Typically, for a single algorithm a single ground-truth is used.
35,961
illuminant performance review evaluation algorithms. estimation of the range then a of We
We then review the performance evaluation of a range of illuminant estimation algorithms.
35,962
images. paper method two-stage proposes the This hyperspectral novel a of classification for
This paper proposes a novel two-stage method for the classification of hyperspectral images.
35,963
information. spatial propose to incorporate approach two-stage In a paper, this we the
In this paper, we propose a two-stage approach to incorporate the spatial information.
35,964
probability will class for be map The resulting noisy. each
The resulting probability map for each class will be noisy.
35,965
information utilizes effectively both data spatial hyperspectral method and sets. spectral of the proposed Our
Our proposed method effectively utilizes both spectral and spatial information of the hyperspectral data sets.
35,966
labelled in a Depth "Relative depths. (RDIS) densely We dataset introduce new Stereo" with relative
We introduce a new "Relative Depth in Stereo" (RDIS) dataset densely labelled with relative depths.
35,967
dataset. RDIS pretrain ResNet on We model first our a
We first pretrain a ResNet model on our RDIS dataset.
35,968
Then we ground-truth on depths. the datasets RGB-D finetune with model metric
Then we finetune the model on RGB-D datasets with metric ground-truth depths.
35,969
During we estimation depth classification task. our formulate a as finetuning,
During our finetuning, we formulate depth estimation as a classification task.
35,970
example-mining concept exploring Instead we of of hard constraint. strategy, large margin introduce the
Instead of exploring hard example-mining strategy, we introduce the concept of large margin constraint.
35,971
two-center on a database organ for of task the We MR experiments perform segmentation.
We perform experiments on a two-center MR database for the task of organ segmentation.
35,972
functioning fundamental Weber's describes of perceptual systems. the law
Weber's law describes the fundamental functioning of perceptual systems.
35,973
pruning. and characteristic important is to an measure ultimately control with It therefore to
It is therefore an important characteristic to measure and ultimately to control with pruning.
35,974
and point individual measured using represented clouds. by The of LiDAR trees was geometry
The geometry of individual trees was measured using LiDAR and represented by point clouds.
35,975
was distribution of discrete sky energy A synthesised public the hemispherical model weather records. using
A discrete energy distribution model of the hemispherical sky was synthesised using public weather records.
35,976
is target domain unsupervised, any domain labels. using Our adaptation without procedure
Our domain adaptation procedure is unsupervised, without using any target domain labels.
35,977
method public validated different chest on segmentation. datasets left/right for our X-ray We two lung
We validated our method on two different chest X-ray public datasets for left/right lung segmentation.
35,978
semantic difficult computer been always segmentation task vision. has a Indoor in
Indoor semantic segmentation has always been a difficult task in computer vision.
35,979
computationally for presents This paper methodology density estimation. efficient new kernel
This paper presents new methodology for computationally efficient kernel density estimation.
35,980
compute density same exactly. estimates to be can The derivative used methodology
The same methodology can be used to compute density derivative estimates exactly.
35,981
is an the size. Given linear computational the in complexity ordered sample sample
Given an ordered sample the computational complexity is linear in the sample size.
35,982
methods in fast the approximation proposed with density results existing extremely Combining methodology estimation.
Combining the proposed methodology with existing approximation methods results in extremely fast density estimation.
35,983
the compared Extensive documents with this approach state-of-the-art. experimentation existing efficiency the and effectiveness of
Extensive experimentation documents the effectiveness and efficiency of this approach compared with the existing state-of-the-art.
35,984
problem, use propose to translation image-to-image models. this we address To
To address this problem, we propose to use image-to-image translation models.
35,985
image both unpaired explore the purpose. We translation of paired this for models usage and
We explore both the usage of paired and unpaired image translation models for this purpose.
35,986
data Combining improvement. to leads further both sources
Combining both data sources leads to further improvement.
35,987
selected is nearest The reference classifier used with the neighbour set.
The nearest neighbour classifier is used with the selected reference set.
35,988
a model. as categorisation can a interpreted be set Such data-driven
Such a set can be interpreted as a data-driven categorisation model.
35,989
cross-referencing. the two models to juxtapose We from the fields enable
We juxtapose the models from the two fields to enable cross-referencing.
35,990
models (VQA) questions open-ended question language answering natural about respond images. to Visual
Visual question answering (VQA) models respond to open-ended natural language questions about images.
35,991
buildings. of derived integrating GBI by is saliency the geometric measured The resulting
The resulting GBI is measured by integrating the derived geometric saliency of buildings.
35,992
degrade face eyeglasses accuracy. significantly recognition of In recognition, the could the application
In the application of face recognition, eyeglasses could significantly degrade the recognition accuracy.
35,993
we a In paper, propose this to synthesis address manner. virtual problem in this
In this paper, we propose to address this problem in a virtual synthesis manner.
35,994
networks actual Convolutional are neural and (ConvNets) image standard classification. recognizement the for
Convolutional neural networks (ConvNets) are the actual standard for image recognizement and classification.
35,995
SqueezeNet, and AlexNet, models study on ConvNets Inception. platform: PyTorch ResNet uses the The available
The study uses ConvNets models available on the PyTorch platform: AlexNet, SqueezeNet, ResNet and Inception.
35,996
goal. of advances approaching the Some recent that we demonstrate are
Some of the recent advances demonstrate we are approaching that goal.
35,997
release research. also our promote We further to code
We also release our code to promote further research.
35,998
rewards. find require agents random in Traditional actions to RL to perform methods exploration
Traditional exploration methods in RL require agents to perform random actions to find rewards.
35,999