2020.02.24.txt

==========New Papers==========

1, TITLE:       On the impressive performance of randomly weighted encoders in summarization tasks

http://arxiv.org/abs/2002.09084

AUTHORS:        Jonathan Pilault ; Jaehong Park ; Christopher Pal

COMMENTS:       Accepted to ACL 2019 SRW. First two authors contributed equally

HIGHLIGHT:      In this work, we investigate the performance of 
untrained randomly initialized encoders in a general class of sequence 
to sequence models and compare their performance with that of 
fully-trained encoders on the task of abstractive summarization.



2, TITLE:       Geometric rank of tensors and subrank of matrix multiplication

http://arxiv.org/abs/2002.09472

AUTHORS:        Swastik Kopparty ; Guy Moshkovitz ; Jeroen Zuiddam

HIGHLIGHT:      Motivated by problems in algebraic complexity theory 
(e.g., matrix multiplication) and extremal combinatorics (e.g., the cap 
set problem and the sunflower problem), we introduce the geometric rank 
as a new tool in the study of tensors and hypergraphs.



3, TITLE:       Temporal Constraint Satisfaction Problems in Fixed-Point Logic

http://arxiv.org/abs/2002.09451

AUTHORS:        Manuel Bodirsky ; Wied Pakusa ; Jakub Rydval

HIGHLIGHT:      We prove that there is no Maltsev condition that 
characterises Datalog already for the CSPs of first-order reducts of 
(Q;<); such CSPs are called temporal CSPs and are of fundamental 
importance in infinite-domain constraint satisfaction.



4, TITLE:       A Road Map to Strong Intelligence

http://arxiv.org/abs/2002.09044

AUTHORS:        Philip Paquette

HIGHLIGHT:      Section 2 proposes a method to classify intelligence, 
and describes the differences between weak and strong intelligence.



5, TITLE:       Learning Dynamic Knowledge Graphs to Generalize on Text-Based Games

http://arxiv.org/abs/2002.09127

AUTHORS:        Ashutosh Adhikari ; Xingdi Yuan ; Marc-Alexandre Côté ; 
Mikuláš Zelinka ; Marc-Antoine Rondeau ; Romain Laroche ; Pascal Poupart
 ; Jian Tang ; Adam Trischler ; William L. Hamilton

HIGHLIGHT:      In this work, we investigate how structured information 
in the form of a knowledge graph (KG) can facilitate effective planning 
and generalization. To train this model to build useful graph 
representations, we introduce and analyze a set of graph-related 
pre-training tasks.



6, TITLE:       Refinement of Unsupervised Cross-Lingual Word Embeddings

http://arxiv.org/abs/2002.09213

AUTHORS:        Magdalena Biesialska ; Marta R. Costa-jussà

COMMENTS:       Accepted at the 24th European Conference on Artificial Intelligence (ECAI 2020)

HIGHLIGHT:      In this paper, we propose a self-supervised method to 
refine the alignment of unsupervised bilingual word embeddings.



7, TITLE:       Guider l'attention dans les modeles de sequence a sequence pour la prediction des actes de dialogue

http://arxiv.org/abs/2002.09419

AUTHORS:        Pierre Colombo ; Emile Chapuis ; Matteo Manica ; Emmanuel Vignon ; Giovanna Varni ; Chloe Clavel

COMMENTS:       in French

HIGHLIGHT:      In this work, we introduce a seq2seq model tailored for 
DA classification using: a hierarchical encoder, a novel guided 
attention mechanism and beam search applied to both training and 
inference.



8, TITLE:       Is Aligning Embedding Spaces a Challenging Task? An Analysis of the Existing Methods

http://arxiv.org/abs/2002.09247

AUTHORS:        Russa Biswas ; Mehwish Alam ; Harald Sack

HIGHLIGHT:      This paper provides a theoretical analysis and 
comparison of the state-of-the-art alignment methods between two 
embedding spaces representing entity-entity and entity-word.



9, TITLE:       The Automated Inspection of Opaque Liquid Vaccines

http://arxiv.org/abs/2002.09406

AUTHORS:        Gregory Palmer ; Benjamin Schnieders ; Rahul Savani ; Karl Tuyls ; Joscha-David Fossel ; Harry Flore

COMMENTS:       8 pages, 5 Figures, 3 Tables, ECAI 2020 Conference Proceedings

HIGHLIGHT:      Using Frame-Completion Generative Adversarial Networks 
we: (i) introduce an algorithm for computing saliency maps, which we use
 to verify that the 3D-ConvNets are indeed identifying anomalies; (ii) 
propose a novel self-training approach using the saliency maps to 
determine if multiple networks agree on the location of anomalies.



10, TITLE:      Fine-Grained Instance-Level Sketch-Based Video Retrieval

http://arxiv.org/abs/2002.09461

AUTHORS:        Peng Xu ; Kun Liu ; Tao Xiang ; Timothy M. Hospedales ; Zhanyu Ma ; Jun Guo ; Yi-Zhe Song

HIGHLIGHT:      In this work, we propose a novel cross-modal retrieval 
problem of fine-grained instance-level sketch-based video retrieval 
(FG-SBVR), where a sketch sequence is used as a query to retrieve a 
specific target video instance.



11, TITLE:      Language as a Cognitive Tool to Imagine Goals in Curiosity-Driven Exploration

http://arxiv.org/abs/2002.09253

AUTHORS:        Cédric Colas ; Tristan Karch ; Nicolas Lair ; 
Jean-Michel Dussoux ; Clément Moulin-Frier ; Peter Ford Dominey ; 
Pierre-Yves Oudeyer

COMMENTS:       Contains main article and supplementaries

HIGHLIGHT:      We present experiments in a simulated domain where the 
agent interacts with procedurally generated scenes containing objects of
 various types and colors, discovers goals, imagines others and learns 
to achieve them.



12, TITLE:      An Advance on Variable Elimination with Applications to Tensor-Based Computation

http://arxiv.org/abs/2002.09320

AUTHORS:        Adnan Darwiche

COMMENTS:       To Appear in Proceedings of the European Conference on Artificial Intelligence (ECAI), Spain, 2020

HIGHLIGHT:      We present new results on the classical algorithm of 
variable elimination, which underlies many algorithms including for 
probabilistic inference.



13, TITLE:      On The Reasons Behind Decisions

http://arxiv.org/abs/2002.09284

AUTHORS:        Adnan Darwiche ; Auguste Hirth

COMMENTS:       To appear in the proceedings of European Conference on Artificial Intelligence (ECAI), Spain, 2020

HIGHLIGHT:      We present a theory for unveiling the reasons behind the
 decisions made by Boolean classifiers and study some of its theoretical
 and practical implications.



14, TITLE:      Detection and Classification of Astronomical Targets 
with Deep Neural Networks in Wide Field Small Aperture Telescopes

http://arxiv.org/abs/2002.09211

AUTHORS:        Peng Jia ; Qiang Liu ; Yongyang Sun

COMMENTS:       Submitted to AAS journal. The complete code can be downloaded from https://github.com/E-Dreamer-LQ/Astronomical_Target_Detection.
 This code can be directly used to process images obtained by WFSATs. 
Images obtained by ordinary sky survey telescopes can also be processed 
with this code, however more annotated images are required to train the 
neural network

HIGHLIGHT:      In this paper, we propose an astronomical targets 
detection and classification framework based on deep neural networks for
 images obtained by wide field small aperture telescopes.



15, TITLE:      Robust Iris Presentation Attack Detection Fusing 2D and 3D Information

http://arxiv.org/abs/2002.09137

AUTHORS:        Zhaoyuan Fang ; Adam Czajka ; Kevin W. Bowyer

HIGHLIGHT:      This paper proposes a method that combines 
two-dimensional and three-dimensional properties of the observed iris to
 address the problem of spoof detection in case when some properties of 
artifacts are unknown.



16, TITLE:      Unsupervised Enhancement of Soft-biometric Privacy with Negative Face Recognition

http://arxiv.org/abs/2002.09181

AUTHORS:        Philipp Terhörst ; Marco Huber ; Naser Damer ; Florian Kirchbuchner ; Arjan Kuijper

COMMENTS:       Currently under review

HIGHLIGHT:      In this work, we present Negative Face Recognition 
(NFR), a novel face recognition approach that enhances the 
soft-biometric privacy on the template-level by representing face 
templates in a complementary (negative) domain.



17, TITLE:      Complete Endomorphisms in Computer Vision

http://arxiv.org/abs/2002.09003

AUTHORS:        Javier Finat ; Francisco Delgado-del-Hoyo

COMMENTS:       22 pages, 2 figures

HIGHLIGHT:      To include these degenerate situations, this paper 
introduces a completion of bilinear maps between spaces given by an 
equivariant compactification of regular transformations.



18, TITLE:      Learning to Inpaint by Progressively Growing the Mask Regions

http://arxiv.org/abs/2002.09280

AUTHORS:        Mohamed Abbas Hedjazi ; Yakup Genc

COMMENTS:       ICCV Workshop on Should we preregister experiments in computer vision?, Seoul, South Korea, 2019

HIGHLIGHT:      This work introduces a new curriculum-style training approach in the context of image inpainting.



19, TITLE:      Stochastic Latent Residual Video Prediction

http://arxiv.org/abs/2002.09219

AUTHORS:        Jean-Yves Franceschi ; Edouard Delasalles ; Mickaël Chen ; Sylvain Lamprier ; Patrick Gallinari

HIGHLIGHT:      In this paper, we overcome these difficulties by 
introducing a novel stochastic temporal model whose dynamics are 
governed in a latent space by a residual update rule.



20, TITLE:      Unsupervised Pre-trained, Texture Aware And Lightweight 
Model for Deep Learning-Based Iris Recognition Under Limited Annotated 
Data

http://arxiv.org/abs/2002.09048

AUTHORS:        Manashi Chakraborty ; Mayukh Roy ; Prabir Kumar Biswas ; Pabitra Mitra

COMMENTS:       Under review at ICIP2020

HIGHLIGHT:      In this paper, we present a texture aware lightweight deep learning framework for iris recognition.



21, TITLE:      Adapted Center and Scale Prediction: More Stable and More Accurate

http://arxiv.org/abs/2002.09053

AUTHORS:        Wenhao Wang

COMMENTS:       13 pages, 7 figures

HIGHLIGHT:      The main contributions of our paper are: (1) We improve the robustness of CSP and make it easier to train.



22, TITLE:      3D U-Net for Segmentation of Plant Root MRI Images in Super-Resolution

http://arxiv.org/abs/2002.09317

AUTHORS:        Yi Zhao ; Nils Wandel ; Magdalena Landl ; Andrea Schnepf ; Sven Behnke

COMMENTS:       6 pages, 5 figures, in the 28th European Symposium on Artificial Neural Networks

HIGHLIGHT:      We propose to increase signal-to-noise ratio and 
resolution by segmenting the scanned volumes into root and soil in 
super-resolution using a 3D U-Net.



23, TITLE:      Brain Age Estimation Using LSTM on Children's Brain MRI

http://arxiv.org/abs/2002.09045

AUTHORS:        Sheng He ; Randy L. Gollub ; Shawn N. Murphy ; Juan 
David Perez ; Sanjay Prabhu ; Rudolph Pienaar ; Richard L. Robertson ; 
P. Ellen Grant ; Yangming Ou

COMMENTS:       ISBI 2020

HIGHLIGHT:      In this paper, we consider the 3D brain MRI volume as a 
sequence of 2D images and propose a new framework using the recurrent 
neural network for brain age estimation.



24, TITLE:      Audio-video Emotion Recognition in the Wild using Deep Hybrid Networks

http://arxiv.org/abs/2002.09023

AUTHORS:        Xin Guo ; Luisa F. Polanía ; Kenneth E. Barner

HIGHLIGHT:      This paper presents an audiovisual-based emotion recognition hybrid network.



25, TITLE:      Affective Expression Analysis in-the-wild using Multi-Task Temporal Statistical Deep Learning Model

http://arxiv.org/abs/2002.09120

AUTHORS:        Nhu-Tai Do ; Soo-Hyung Kim

HIGHLIGHT:      In this paper, we present an affective expression analysis model that deals with the above challenges.



26, TITLE:      Learning Fairness-aware Relational Structures

http://arxiv.org/abs/2002.09471

AUTHORS:        Yue Zhang ; Arti Ramesh

COMMENTS:       Accepted for publication in ECAI 2020

HIGHLIGHT:      In this work, we introduce Fair-A3SL, a fairness-aware 
structure learning algorithm for learning relational structures, which 
incorporates fairness measures while learning relational graphical model
 structures.



27, TITLE:      Accelerating Reinforcement Learning with a Directional-Gaussian-Smoothing Evolution Strategy

http://arxiv.org/abs/2002.09077

AUTHORS:        Jiaxing Zhang ; Hoang Tran ; Guannan Zhang

HIGHLIGHT:      In this work, we employ a Directional Gaussian Smoothing
 Evolutionary Strategy (DGS-ES) to accelerate RL training, which is 
well-suited to address these two challenges with its ability to i) 
provide gradient estimates with high accuracy, and ii) find nonlocal 
search direction which lays stress on large-scale variation of the 
reward function and disregards local fluctuation.



28, TITLE:      Convolutional Tensor-Train LSTM for Spatio-temporal Learning

http://arxiv.org/abs/2002.09131

AUTHORS:        Jiahao Su ; Wonmin Byeon ; Furong Huang ; Jan Kautz ; Animashree Anandkumar

COMMENTS:       Jiahao Su and Wonmin Byeon contributed equally to this work. 17 pages, 14 figures

HIGHLIGHT:      To address this issue, we propose Convolutional 
Tensor-Train Decomposition (CTTD), a novel tensor decomposition with 
convolutional operations.



29, TITLE:      SemanticPOSS: A Point Cloud Dataset with Large Quantity of Dynamic Instances

http://arxiv.org/abs/2002.09147

AUTHORS:        Yancheng Pan ; Biao Gao ; Jilin Mei ; Sibo Geng ; Chengkun Li ; Huijing Zhao

COMMENTS:       submited to IEEE Intelligent Vehicles Symposium(2020)

HIGHLIGHT:      In this paper, we propose the SemanticPOSS dataset, 
which contains 2988 various and complicated LiDAR scans with large 
quantity of dynamic instances.



30, TITLE:      Greedy Policy Search: A Simple Baseline for Learnable Test-Time Augmentation

http://arxiv.org/abs/2002.09103

AUTHORS:        Dmitry Molchanov ; Alexander Lyzhov ; Yuliya Molchanova ; Arsenii Ashukha ; Dmitry Vetrov

HIGHLIGHT:      The primary goal of this paper is to demonstrate that 
test-time augmentation policies can be successfully learned too.



31, TITLE:      BlockGAN: Learning 3D Object-aware Scene Representations from Unlabelled Images

http://arxiv.org/abs/2002.08988

AUTHORS:        Thu Nguyen-Phuoc ; Christian Richardt ; Long Mai ; Yong-Liang Yang ; Niloy Mitra

COMMENTS:       For project page, see https://www.monkeyoverflow.com/#/blockgan/

HIGHLIGHT:      We present BlockGAN, an image generative model that 
learns object-aware 3D scene representations directly from unlabelled 2D
 images.



32, TITLE:      A Hybrid Algorithm Based Robust Big Data Clustering for 
Solving Unhealthy Initialization, Dynamic Centroid Selection and Empty 
clustering Problems with Analysis

http://arxiv.org/abs/2002.09380

AUTHORS:        Y. A. Joarder ; Mosabbir Ahmed

COMMENTS:       18 Pages

HIGHLIGHT:      Big Data is a massive volume of both structured and 
unstructured data that is too large and it also difficult to process 
using traditional techniques.



33, TITLE:      It's Not What Machines Can Learn, It's What We Cannot Teach

http://arxiv.org/abs/2002.09398

AUTHORS:        Gal Yehuda ; Moshe Gabel ; Assaf Schuster

HIGHLIGHT:      In this work we offer a different perspective on this question.



34, TITLE:      Accessing Higher-level Representations in Sequential Transformers with Feedback Memory

http://arxiv.org/abs/2002.09402

AUTHORS:        Angela Fan ; Thibaut Lavril ; Edouard Grave ; Armand Joulin ; Sainbayar Sukhbaatar

HIGHLIGHT:      In this work, we propose the Feedback Transformer 
architecture that exposes all previous representations to all future 
representations, meaning the lowest representation of the current 
timestep is formed from the highest-level abstract representation of the
 past.



35, TITLE:      How to Evaluate Solutions in Pareto-based Search-Based 
Software Engineering? A Critical Review and Methodological Guidance

http://arxiv.org/abs/2002.09040

AUTHORS:        Tao Chen ; Miqing Li ; Xin Yao

COMMENTS:       submitted, 7 figures and 5 tables

HIGHLIGHT:      In this paper, we carry out a critical review of quality evaluation for multiobjective optimization in SBSE.



36, TITLE:      Learning Precise 3D Manipulation from Multiple Uncalibrated Cameras

http://arxiv.org/abs/2002.09107

AUTHORS:        Iretiayo Akinola ; Jacob Varley ; Dmitry Kalashnikov

COMMENTS:       Accepted at International Conference on Robotics and Automation (ICRA 2020)

HIGHLIGHT:      In this work, we present an effective multi-view 
approach to closed-loop end-to-end learning of precise manipulation 
tasks that are 3D in nature.



37, TITLE:      Curating Social Media Data

http://arxiv.org/abs/2002.09202

AUTHORS:        Kushal Vaghani

COMMENTS:       Masters by Research Thesis

HIGHLIGHT:      We propose a data curation pipeline, namely 
CrowdCorrect, to enable analysts cleansing and curating social data and 
preparing it for reliable analytics.



38, TITLE:      Leveraging Photogrammetric Mesh Models for Aerial-Ground
 Feature Point Matching Toward Integrated 3D Reconstruction

http://arxiv.org/abs/2002.09085

AUTHORS:        Qing Zhu ; Zhendong Wang ; Han Hu ; Linfu Xie ; Xuming Ge ; Yeting Zhang

HIGHLIGHT:      To address these problems, we propose a novel approach: 
leveraging photogrammetric mesh models for aerial-ground image matching.



39, TITLE:      Crowdsourced Collective Entity Resolution with Relational Match Propagation

http://arxiv.org/abs/2002.09361

AUTHORS:        Jiacheng Huang ; Wei Hu ; Zhifeng Bao ; Yuzhong Qu

COMMENTS:       Accepted by the 36th IEEE International Conference on Data Engineering (ICDE 2020)

HIGHLIGHT:      In this paper, we propose a novel approach called 
crowdsourced collective ER, which leverages the relationships between 
entities to infer matches jointly rather than independently.



40, TITLE:      Total tessellation cover and quantum walk

http://arxiv.org/abs/2002.08992

AUTHORS:        Alexandre Abreu ; Luís Cunha ; Celina de Figueiredo ; Franklin Marquezino ; Daniel Posner ; Renato Portugal

HIGHLIGHT:      We propose the total staggered quantum walk model and the total tessellation cover of a graph.



41, TITLE:      Comparing Different Deep Learning Architectures for Classification of Chest Radiographs

http://arxiv.org/abs/2002.08991

AUTHORS:        Keno K. Bressem ; Lisa Adams ; Christoph Erxleben ; Bernd Hamm ; Stefan Niehues ; Janis Vahldiek

COMMENTS:       15 pages, 6 figures, 3 tables

HIGHLIGHT:      Chest radiographs are among the most frequently acquired
 images in radiology and are often the subject of computer vision 
research.



42, TITLE:      Comparing recurrent and convolutional neural networks for predicting wave propagation

http://arxiv.org/abs/2002.08981

AUTHORS:        Stathi Fotiadis ; Eduardo Pignatelli ; Mario Lino Valencia ; Chris Cantwell ; Amos Storkey ; Anil A. Bharath

HIGHLIGHT:      In this work, we investigate the performance of 
recurrent and convolutional deep neural network architectures to predict
 the surface waves.



43, TITLE:      Affinity and Diversity: Quantifying Mechanisms of Data Augmentation

http://arxiv.org/abs/2002.08973

AUTHORS:        Raphael Gontijo-Lopes ; Sylvia J. Smullin ; Ekin D. Cubuk ; Ethan Dyer

COMMENTS:       10 pages, 7 figures

HIGHLIGHT:      To this end, we introduce interpretable and easy-to-compute measures: Affinity and Diversity.



44, TITLE:      Anonymizing Data for Privacy-Preserving Federated Learning

http://arxiv.org/abs/2002.09096

AUTHORS:        Olivia Choudhury ; Aris Gkoulalas-Divanis ; Theodoros 
Salonidis ; Issa Sylla ; Yoonyoung Park ; Grace Hsu ; Amar Das

COMMENTS:       24th European Conference on Artificial Intelligence (ECAI)

HIGHLIGHT:      In this paper, we propose the first syntactic approach 
for offering privacy in the context of federated learning.



45, TITLE:      Face Phylogeny Tree Using Basis Functions

http://arxiv.org/abs/2002.09068

AUTHORS:        Sudipta Banerjee ; Arun Ross

HIGHLIGHT:      In this work, we utilize three different families of 
basis functions to model pairwise relationships between near-duplicate 
images.



46, TITLE:      Residual Knowledge Distillation

http://arxiv.org/abs/2002.09168

AUTHORS:        Mengya Gao ; Yujun Shen ; Quanquan Li ; Chen Change Loy

COMMENTS:       9 pages, 3 figures, 3 tables

HIGHLIGHT:      In this way, S and A complement with each other to get 
better knowledge from T. Furthermore, we devise an effective method to 
derive S and A from a given model without increasing the total 
computational cost.



47, TITLE:      Disentangling Controllable Object through Video Prediction Improves Visual Reinforcement Learning

http://arxiv.org/abs/2002.09136

AUTHORS:        Yuanyi Zhong ; Alexander Schwing ; Jian Peng

COMMENTS:       Accepted to ICASSP 2020

HIGHLIGHT:      Leveraging action-conditioned video prediction, we 
propose an end-to-end learning framework to disentangle the controllable
 object from the observation signal.



48, TITLE:      Bidirectional Generative Modeling Using Adversarial Gradient Estimation

http://arxiv.org/abs/2002.09161

AUTHORS:        Xinwei Shen ; Tong Zhang ; Kani Chen

HIGHLIGHT:      We present a new optimization method for this 
formulation, where the gradient is computed using an adversarially 
learned discriminator.



49, TITLE:      Real-Time Optimal Guidance and Control for Interplanetary Transfers Using Deep Networks

http://arxiv.org/abs/2002.09063

AUTHORS:        Dario Izzo ; Ekin Öztürk

HIGHLIGHT:      We consider the Earth-Venus mass-optimal interplanetary 
transfer of a low-thrust spacecraft and show how the optimal guidance 
can be represented by deep networks in a large portion of the state 
space and to a high degree of accuracy.



50, TITLE:      An Evolutionary Deep Learning Method for Short-term Wind
 Speed Prediction: A Case Study of the Lillgrund Offshore Wind Farm

http://arxiv.org/abs/2002.09106

AUTHORS:        Mehdi Neshat ; Meysam Majidi Nezhad ; Ehsan Abbasnejad ;
 Lina Bertling Tjernberg ; Davide Astiaso Garcia ; Bradley Alexander ; 
Markus Wagner

HIGHLIGHT:      This study uses a new hybrid evolutionary approach that 
uses a popular evolutionary search algorithm, CMA-ES, to tune the 
hyper-parameters of two Long short-term memory(LSTM) ANN models for wind
 prediction.



51, TITLE:      Recent Trends in the Use of Statistical Tests for 
Comparing Swarm and Evolutionary Computing Algorithms: Practical 
Guidelines and a Critical Review

http://arxiv.org/abs/2002.09227

AUTHORS:        J. Carrasco ; S. García ; M. M. Rueda ; S. Das ; F. Herrera

COMMENTS:       52 pages, 10 figures, 19 tables

HIGHLIGHT:      In the present paper we gather and examine the 
approaches taken from different perspectives to summarise the 
assumptions made by these statistical tests, the conclusions reached and
 the steps followed to perform them correctly.



52, TITLE:      Symbolic Execution Game Semantics

http://arxiv.org/abs/2002.09115

AUTHORS:        Yu-Yang Lin ; Nikos Tzevelekos

COMMENTS:       41 pages, 5 figures

HIGHLIGHT:      We present a framework for symbolically executing and 
model checking higher-order programs with external (open) methods.



53, TITLE:      RustHorn: CHC-based Verification for Rust Programs (full version)

http://arxiv.org/abs/2002.09002

AUTHORS:        Yusuke Matsushita ; Takeshi Tsukada ; Naoki Kobayashi

COMMENTS:       Full version of the same-titled paper in ESOP2020

HIGHLIGHT:      This paper proposes a novel translation of 
pointer-manipulating Rust programs into CHCs, which clears away pointers
 and heaps by leveraging ownership.



54, TITLE:      Graph4Code: A Machine Interpretable Knowledge Graph for Code

http://arxiv.org/abs/2002.09440

AUTHORS:        Kavitha Srinivas ; Ibrahim Abdelaziz ; Julian Dolby ; James P. McCusker

HIGHLIGHT:      We use the Whyis knowledge graph management framework to make the graph easily extensible.



55, TITLE:      Calibrating Deep Neural Networks using Focal Loss

http://arxiv.org/abs/2002.09437

AUTHORS:        Jishnu Mukhoti ; Viveka Kulharia ; Amartya Sanyal ; Stuart Golodetz ; Philip H. S. Torr ; Puneet K. Dokania

HIGHLIGHT:      We provide a thorough analysis of the factors causing 
miscalibration, and use the insights we glean from this to justify the 
empirically excellent performance of focal loss.



56, TITLE:      Post-training Quantization with Multiple Points: Mixed Precision without Mixed Precision

http://arxiv.org/abs/2002.09049

AUTHORS:        Xingchao Liu ; Mao Ye ; Dengyong Zhou ; Qiang Liu

HIGHLIGHT:      We propose multipoint quantization, a quantization 
method that approximates a full-precision weight vector using a linear 
combination of multiple vectors of low-bit numbers; this is in contrast 
to typical quantization methods that approximate each weight using a 
single low precision number.



57, TITLE:      Learning Intermediate Features of Object Affordances with a Convolutional Neural Network

http://arxiv.org/abs/2002.08975

AUTHORS:        Aria Yuan Wang ; Michael J. Tarr

COMMENTS:       Published on 2018 Conference on Cognitive Computational Neuroscience. See <https://ccneuro.org/2018/Papers/ViewPapers.asp?PaperNum=1134>

HIGHLIGHT:      To better understand the nature of affordances, we 
trained a deep convolutional neural network (CNN) to recognize 
affordances from images and to learn the underlying features or the 
dimensionality of affordances.



58, TITLE:      Designing Fair AI for Managing Employees in Organizations: A Review, Critique, and Design Agenda

http://arxiv.org/abs/2002.09054

AUTHORS:        Lionel P. Robert ; Casey Pierce ; Liz Morris ; Sangmi Kim ; Rasha Alahmad

COMMENTS:       66 pages, 2 figures

HIGHLIGHT:      This paper addresses the issue in three ways.



59, TITLE:      Exploiting the Full Capacity of Deep Neural Networks 
while Avoiding Overfitting by Targeted Sparsity Regularization

http://arxiv.org/abs/2002.09237

AUTHORS:        Karim Huesmann ; Soeren Klemm ; Lars Linsen ; Benjamin Risse

COMMENTS:       10 pages, 9 figures

HIGHLIGHT:      Here, we demonstrate that neural network activation 
sparsity is a reliable indicator for overfitting which we utilize to 
propose novel targeted sparsity visualization and regularization 
strategies.



60, TITLE:      Efficient Learning of Model Weights via Changing Features During Training

http://arxiv.org/abs/2002.09249

AUTHORS:        Marcell Beregi-Kovács ; Ágnes Baran ; András Hajdu

HIGHLIGHT:      In this paper, we propose a machine learning model, which dynamically changes the features during training.



61, TITLE:      Binary Probability Model for Learning Based Image Compression

http://arxiv.org/abs/2002.09259

AUTHORS:        Théo Ladune ; Pierrick Philippe ; Wassim Hamidouche ; Lu Zhang ; Olivier Deforges

HIGHLIGHT:      In this paper, we propose to enhance learned image 
compression systems with a richer probability model for the latent 
variables.







==========Updates to Previous Papers==========

1, TITLE:       Groups with ALOGTIME-hard word problems and PSPACE-complete compressed word problems

http://arxiv.org/abs/1909.13781

AUTHORS:        Laurent Bartholdi ; Michael Figelius ; Markus Lohrey ; Armin Weiß

HIGHLIGHT:      We give lower bounds on the complexity of the word 
problem of certain non-solvable groups: for a large class of 
non-solvable infinite groups, including in particular free groups, 
Grigorchuk's group and Thompson's groups, we prove that their word 
problem is $\mathsf{NC}^1$-hard.



2, TITLE:       Combinatorial Characterisations of Graphical Computational Search Problems

http://arxiv.org/abs/1802.03028

AUTHORS:        Koko Kalambay Kayibi

HIGHLIGHT:      A Graphical Search problem, denoted $\Pi(X,\gamma)$, 
where $X$ is the vertex set or edge set of a graph $G$, consists of 
finding a solution $Y$, where $Y \subseteq X$ and $Y$ satisfies the 
predicate $\gamma$.



3, TITLE:       `Why not give this work to them?' Explaining AI-Moderated Task-Allocation Outcomes using Negotiation Trees

http://arxiv.org/abs/2002.01640

AUTHORS:        Zahra Zahedi ; Sailik Sengupta ; Subbarao Kambhampati

COMMENTS:       First two authors are equal contributors

HIGHLIGHT:      To address these challenges, we introduce the notion of 
an AI Task Allocator (AITA) that, with complete knowledge, comes up with
 fair allocations that strike a balance between the individual human 
costs and the team's performance cost.



4, TITLE:       Multilingual acoustic word embedding models for processing zero-resource languages

http://arxiv.org/abs/2002.02109

AUTHORS:        Herman Kamper ; Yevgen Matusevych ; Sharon Goldwater

COMMENTS:       5 pages, 4 figures, 1 table; accepted to ICASSP 2020. arXiv admin note: text overlap with arXiv:1811.00403

HIGHLIGHT:      Here we propose to train a single supervised embedding 
model on labelled data from multiple well-resourced languages and then 
apply it to unseen zero-resource languages.



5, TITLE:       Most Important Fundamental Rule of Poker Strategy

http://arxiv.org/abs/1906.09895

AUTHORS:        Sam Ganzfried ; Max Chiswick

HIGHLIGHT:      A recent line of research has explored approaches for 
extrapolating knowledge from strong game-theoretic strategies that can 
be understood by humans.



6, TITLE:       A One-to-One Correspondence between Natural Numbers and Binary Trees

http://arxiv.org/abs/2002.04477

AUTHORS:        Osvaldo Skliar ; Sherry Gapper ; Ricardo E. Monge

COMMENTS:       30 pages

HIGHLIGHT:      A One-to-One Correspondence between Natural Numbers and Binary Trees



7, TITLE:       The Open Images Dataset V4: Unified image 
classification, object detection, and visual relationship detection at 
scale

http://arxiv.org/abs/1811.00982

AUTHORS:        Alina Kuznetsova ; Hassan Rom ; Neil Alldrin ; Jasper 
Uijlings ; Ivan Krasin ; Jordi Pont-Tuset ; Shahab Kamali ; Stefan Popov
 ; Matteo Malloci ; Alexander Kolesnikov ; Tom Duerig ; Vittorio Ferrari

COMMENTS:       Accepted to IJCV

HIGHLIGHT:      We present Open Images V4, a dataset of 9.2M images with
 unified annotations for image classification, object detection and 
visual relationship detection.



8, TITLE:       Automatic Shortcut Removal for Self-Supervised Representation Learning

http://arxiv.org/abs/2002.08822

AUTHORS:        Matthias Minderer ; Olivier Bachem ; Neil Houlsby ; Michael Tschannen

HIGHLIGHT:      Here, we propose a general framework for removing shortcut features automatically.



9, TITLE:       Adaptive Quantile Sparse Image (AQuaSI) Prior for Inverse Imaging Problems

http://arxiv.org/abs/1804.02152

AUTHORS:        Franziska Schirrmacher ; Thomas K öhler ; Christian Riess

HIGHLIGHT:      In this work, we propose the Adaptive Quantile Sparse Image (AQuaSI) prior.



10, TITLE:      Parallel Algorithm for Approximating Nash Equilibrium in
 Multiplayer Stochastic Games with Application to Naval Strategic 
Planning

http://arxiv.org/abs/1910.00193

AUTHORS:        Sam Ganzfried ; Conner Laughlin ; Charles Morefield

HIGHLIGHT:      We present a new algorithm for these settings, which 
constitutes the first parallel algorithm for multiplayer stochastic 
games.



11, TITLE:      MDE: Multiple Distance Embeddings for Link Prediction in Knowledge Graphs

http://arxiv.org/abs/1905.10702

AUTHORS:        Afshin Sadeghi ; Damien Graux ; Hamed Shariat Yazdi ; Jens Lehmann

COMMENTS:       Accepted paper in ECAI 2020

HIGHLIGHT:      We propose the Multiple Distance Embedding model (MDE) 
that addresses these limitations and a framework to collaboratively 
combine variant latent distance-based terms.



12, TITLE:      Ludii -- The Ludemic General Game System

http://arxiv.org/abs/1905.05013

AUTHORS:        Éric Piette ; Dennis J. N. J. Soemers ; Matthew 
Stephenson ; Chiara F. Sironi ; Mark H. M. Winands ; Cameron Browne

COMMENTS:       Accepted at ECAI 2020

HIGHLIGHT:      In this paper, we describe the "ludemic" general game 
system Ludii, which has the potential to provide an efficient tool for 
AI researchers as well as game designers, historians, educators and 
practitioners in related fields.



13, TITLE:      Accelerating Column Generation via Flexible Dual Optimal Inequalities with Application to Entity Resolution

http://arxiv.org/abs/1909.05460

AUTHORS:        Vishnu Suresh Lokhande ; Shaofei Wang ; Maneesh Singh ; Julian Yarkony

COMMENTS:       Accepted at AAAI20. Version update. Update the link to project page

HIGHLIGHT:      In this paper, we introduce a new optimization approach to Entity Resolution.



14, TITLE:      Semantics of negative sequential patterns

http://arxiv.org/abs/2002.06920

AUTHORS:        Thomas Guyet ; Philippe Besnard

COMMENTS:       proceedings of European Conference on Artificial Intelligence (ECAI), Jun 2020, Santiago de Compostela, Spain

HIGHLIGHT:      These semantics are illustrated and formally studied, in
 particular we propose dominance and equivalence relations between them.



15, TITLE:      Mathematical Foundations for Designing and Development of Intelligent Systems of Information Analysis

http://arxiv.org/abs/1510.04183

AUTHORS:        D. O. Terletskyi ; O. I. Provotar

HIGHLIGHT:      This paper contains analysis of basic concepts of OOP 
and their relation with set theory and artificial intelligence (AI).



16, TITLE:      Complexity of full counting statistics of free quantum particles in product states

http://arxiv.org/abs/1904.06069

AUTHORS:        Dmitri A. Ivanov ; Leonid Gurvits

COMMENTS:       8 pages, published version

HIGHLIGHT:      We study the computational complexity of 
quantum-mechanical expectation values of single-particle operators in 
bosonic and fermionic multi-particle product states.



17, TITLE:      FakeLocator: Robust Localization of GAN-Based Face 
Manipulations via Semantic Segmentation Networks with Bells and Whistles

http://arxiv.org/abs/2001.09598

AUTHORS:        Yihao Huang ; Felix Juefei-Xu ; Run Wang ; Xiaofei Xie ;
 Lei Ma ; Jianwen Li ; Weikai Miao ; Yang Liu ; Geguang Pu

COMMENTS:       7 pages

HIGHLIGHT:      Compared with the baseline, our method performs several times better on various metrics.



18, TITLE:      Structured Knowledge Distillation for Dense Prediction

http://arxiv.org/abs/1903.04197

AUTHORS:        Yifan Liu ; Changyong Shun ; Jingdong Wang ; Chunhua Shen

COMMENTS:       v1:10 pages cvpr2019 accepted; v2:15 pages for a journal version; Code is available at: https://github.com/irfanICMLL/structure_knowledge_distillation; fix typos

HIGHLIGHT:      In this paper, we consider transferring the structure 
information from large networks to small ones for dense prediction 
tasks.



19, TITLE:      Stabilizing DARTS with Amended Gradient Estimation on Architectural Parameters

http://arxiv.org/abs/1910.11831

AUTHORS:        Kaifeng Bi ; Changping Hu ; Lingxi Xie ; Xin Chen ; Longhui Wei ; Qi Tian

COMMENTS:       21 pages, 11 figures, submitted to ICML 2020, extensive results added, updated the 2nd row of Table 5 beyond v3

HIGHLIGHT:      Our approach bridges the gap from two aspects, namely, 
amending the estimation on the architectural gradients, and unifying the
 hyper-parameter settings in the search and re-training stages.



20, TITLE:      Learning Object Placements For Relational Instructions by Hallucinating Scene Representations

http://arxiv.org/abs/2001.08481

AUTHORS:        Oier Mees ; Alp Emek ; Johan Vertens ; Wolfram Burgard

COMMENTS:       Accepted at the 2020 IEEE International Conference on Robotics and Automation (ICRA). Video at https://www.youtube.com/watch?v=zaZkHTWFMKM

HIGHLIGHT:      In this work, we present a convolutional neural network 
for estimating pixelwise object placement probabilities for a set of 
spatial relations from a single input image.



21, TITLE:      An Overview of Perception and Decision-Making in Autonomous Systems in the Era of Learning

http://arxiv.org/abs/2001.02319

AUTHORS:        Yang Tang ; Chaoqiang Zhao ; Jianrui Wang ; Chongzhen 
Zhang ; Qiyu Sun ; Weixing Zheng ; Wenli Du ; Feng Qian ; Juergen kurths

COMMENTS:       27 pages, 8 figures

HIGHLIGHT:      In this review, we focus on the applications of 
learning-based approaches in perception and decision-making in 
autonomous systems, which is different from previous reviews that 
discussed traditional methods.



22, TITLE:      R3Det: Refined Single-Stage Detector with Feature Refinement for Rotating Object

http://arxiv.org/abs/1908.05612

AUTHORS:        Xue Yang ; Qingqing Liu ; Junchi Yan ; Ang Li ; Zhiqiang Zhang ; Gang Yu

COMMENTS:       13 pages, 14 figures, 10 tables

HIGHLIGHT:      In this paper, we propose an end-to-end refined 
single-stage rotation detector for fast and accurate positioning 
objects.



23, TITLE:      NeuRoRA: Neural Robust Rotation Averaging

http://arxiv.org/abs/1912.04485

AUTHORS:        Pulak Purkait ; Tat-Jun Chin ; Ian Reid

HIGHLIGHT:      In this work, we aim to build a neural network that 
learns the noise patterns from the data and predict/regress the model 
parameters from the noisy relative orientations.



24, TITLE:      A Large-scale Study of Representation Learning with the Visual Task Adaptation Benchmark

http://arxiv.org/abs/1910.04867

AUTHORS:        Xiaohua Zhai ; Joan Puigcerver ; Alexander Kolesnikov ; 
Pierre Ruyssen ; Carlos Riquelme ; Mario Lucic ; Josip Djolonga ; Andre 
Susano Pinto ; Maxim Neumann ; Alexey Dosovitskiy ; Lucas Beyer ; 
Olivier Bachem ; Michael Tschannen ; Marcin Michalski ; Olivier Bousquet
 ; Sylvain Gelly ; Neil Houlsby

HIGHLIGHT:      We present the Visual Task Adaptation Benchmark (VTAB), 
which defines good representations as those that adapt to diverse, 
unseen tasks with few examples.



25, TITLE:      Policy Optimization Through Approximate Importance Sampling

http://arxiv.org/abs/1910.03857

AUTHORS:        Marcin B. Tomczak ; Dongho Kim ; Peter Vrancx ; Kee-Eung Kim

HIGHLIGHT:      In this paper we derive an alternative objective that 
obtains the value of the target policy by applying importance sampling 
(IS).



26, TITLE:      Minimalistic Attacks: How Little it Takes to Fool a Deep Reinforcement Learning Policy

http://arxiv.org/abs/1911.03849

AUTHORS:        Xinghua Qu ; Zhu Sun ; Yew-Soon Ong ; Abhishek Gupta ; Pengfei Wei

HIGHLIGHT:      Recent studies have revealed that neural network-based policies can be easily fooled by adversarial examples.



27, TITLE:      Forecaster: A Graph Transformer for Forecasting Spatial and Time-Dependent Data

http://arxiv.org/abs/1909.04019

AUTHORS:        Yang Li ; José M. F. Moura

HIGHLIGHT:      To address these challenges, we propose Forecaster, a graph Transformer architecture.



28, TITLE:      Linking Physicians to Medical Research Results via Knowledge Graph Embeddings and Twitter

http://arxiv.org/abs/1908.02571

AUTHORS:        Afshin Sadeghi ; Jens Lehmann

COMMENTS:       AI for Good, Data Science for Social Good, Machine 
learning for Social Good, Twitter Data, Knowledge Graph Embeddings, 
Medical Research

HIGHLIGHT:      In this study, we apply Multi Distance Knowledge Graph 
Embeddings (MDE) to link physicians and surgeons to the latest medical 
breakthroughs that are shared as the research results on Twitter.



29, TITLE:      Learning Non-Parametric Invariances from Data with Permanent Random Connectomes

http://arxiv.org/abs/1911.05266

AUTHORS:        Dipan K. Pal ; Akshay Chawla ; Marios Savvides

COMMENTS:       Preprint (accepted at NeurIPS SVRHM 2019 Workshop)

HIGHLIGHT:      In this paper, we introduce a new architectural layer 
for convolutional networks which is capable of learning general 
invariances from data itself.



30, TITLE:      Adversarially robust transfer learning

http://arxiv.org/abs/1905.08232

AUTHORS:        Ali Shafahi ; Parsa Saadatpanah ; Chen Zhu ; Amin Ghiasi ; Christoph Studer ; David Jacobs ; Tom Goldstein

HIGHLIGHT:      We start by observing that robust networks contain robust feature extractors.



31, TITLE:      Single Image Super-resolution via Dense Blended Attention Generative Adversarial Network for Clinical Diagnosis

http://arxiv.org/abs/1906.06575

AUTHORS:        Kewen Liu ; Yuan Ma ; Hongxia Xiong ; Zejun Yan ; Zhijun
 Zhou ; Chaoyang Liu ; Panpan Fang ; Xiaojun Li ; Yalei Chen

COMMENTS:       We abandoned this paper due to its limitation only 
applied on medical images, please view our lastest work at 
arXiv:1911.03464

HIGHLIGHT:      In order to address the issue that medical images would 
suffer from severe blurring caused by lack of high-frequency details, 
this paper develops a novel image super-resolution(SR) algorithm called 
SR-DBAN via dense neural network and blended attention mechanism.



32, TITLE:      Program Sketching with Live Bidirectional Evaluation

http://arxiv.org/abs/1911.00583

AUTHORS:        Justin Lubin ; Nick Collins ; Cyrus Omar ; Ravi Chugh

HIGHLIGHT:      We present Sketch-n-Myth, a technique for completing 
program sketches whereby the evaluation of ordinary assertions gives 
rise to input-output examples.



33, TITLE:      Interaction-aware Decision Making with Adaptive Strategies under Merging Scenarios

http://arxiv.org/abs/1904.06025

AUTHORS:        Yeping Hu ; Alireza Nakhaei ; Masayoshi Tomizuka ; Kikuo Fujimura

COMMENTS:       Best Paper Finalist of IROS 2019

HIGHLIGHT:      In this paper, we proposed an interaction-aware decision
 making with adaptive strategies (IDAS) approach that can let the 
autonomous vehicle negotiate the road with other drivers by leveraging 
their cooperativeness under merging scenarios.



34, TITLE:      Differentially Private Meta-Learning

http://arxiv.org/abs/1909.05830

AUTHORS:        Jeffrey Li ; Mikhail Khodak ; Sebastian Caldas ; Ameet Talwalkar

HIGHLIGHT:      Empirically, we apply our analysis to the problems of 
federated learning with personalization and few-shot classification, 
showing that allowing the relaxation to task-global privacy from the 
more commonly studied notion of local privacy leads to dramatically 
increased performance in recurrent neural language modeling and image 
classification.



35, TITLE:      Optimal Transfer Learning Model for Binary Classification of Funduscopic Images through Simple Heuristics

http://arxiv.org/abs/2002.04189

AUTHORS:        Rohit Jammula ; Vishnu Rajan Tejus ; Shreya Shankar

COMMENTS:       5 pages. 4 tables. Accepted to present in Machine Learning in Computational Biology (MLCB) 2019 conference

HIGHLIGHT:      In an effort to broaden the applications of 
computer-aided ocular disease diagnosis, we propose a unifying model for
 disease classification: low-cost inference of a fundus image to 
determine whether it is healthy or diseased.



36, TITLE:      Exploratory Not Explanatory: Counterfactual Analysis of Saliency Maps for Deep Reinforcement Learning

http://arxiv.org/abs/1912.05743

AUTHORS:        Akanksha Atrey ; Kaleigh Clary ; David Jensen

COMMENTS:       Published at ICLR 2020

HIGHLIGHT:      We introduce an empirical approach grounded in 
counterfactual reasoning to test the hypotheses generated from saliency 
maps and assess the degree to which they correspond to the semantics of 
RL environments.



37, TITLE:      Soft Threshold Weight Reparameterization for Learnable Sparsity

http://arxiv.org/abs/2002.03231

AUTHORS:        Aditya Kusupati ; Vivek Ramanujan ; Raghav Somani ; Mitchell Wortsman ; Prateek Jain ; Sham Kakade ; Ali Farhadi

COMMENTS:       17 pages, 9 figures

HIGHLIGHT:      This work proposes Soft Threshold Reparameterization 
(STR), a novel use of the soft-threshold operator on DNN weights.



38, TITLE:      Towards neural networks that provably know when they don't know

http://arxiv.org/abs/1909.12180

AUTHORS:        Alexander Meinke ; Matthias Hein

HIGHLIGHT:      In this paper we propose a new approach to OOD which overcomes both problems.



39, TITLE:      Mixing autoencoder with classifier: conceptual data visualization

http://arxiv.org/abs/1912.01137

AUTHORS:        Pitoyo Hartono

HIGHLIGHT:      In this short paper, a neural network that is able to 
form a low dimensional topological hidden representation is explained.



40, TITLE:      Linear Mode Connectivity and the Lottery Ticket Hypothesis

http://arxiv.org/abs/1912.05671

AUTHORS:        Jonathan Frankle ; Gintare Karolina Dziugaite ; Daniel M. Roy ; Michael Carbin

COMMENTS:       This submission subsumes 1903.01611 ("Stabilizing the 
Lottery Ticket Hypothesis" and "The Lottery Ticket Hypothesis at Scale")

HIGHLIGHT:      We introduce "instability analysis," which assesses 
whether a neural network optimizes to the same, linearly connected 
minimum under different samples of SGD noise.



41, TITLE:      SemI2I: Semantically Consistent Image-to-Image Translation for Domain Adaptation of Remote Sensing Data

http://arxiv.org/abs/2002.05925

AUTHORS:        Onur Tasar ; S L Happy ; Yuliya Tarabalka ; Pierre Alliez

HIGHLIGHT:      To address this issue, we propose a new data 
augmentation approach that transfers the style of test data to training 
data using generative adversarial networks.