February 2, 2020

3348 words 16 mins read

Paper Group AWR 39

Grapy-ML: Graph Pyramid Mutual Learning for Cross-dataset Human Parsing. Computação Urbana da Teoria à Prática: Fundamentos, Aplicações e Desafios. Giveme5W1H: A Universal System for Extracting Main Events from News Articles. Adversarial point perturbations on 3D objects. StructureNet: Hierarchical Graph Networks for 3D Shape Generation. 3DN: 3D De …

Grapy-ML: Graph Pyramid Mutual Learning for Cross-dataset Human Parsing


Title	Grapy-ML: Graph Pyramid Mutual Learning for Cross-dataset Human Parsing
Authors	Haoyu He, Jing Zhang, Qiming Zhang, Dacheng Tao
Abstract	Human parsing, or human body part semantic segmentation, has been an active research topic due to its wide potential applications. In this paper, we propose a novel GRAph PYramid Mutual Learning (Grapy-ML) method to address the cross-dataset human parsing problem, where the annotations are at different granularities. Starting from the prior knowledge of the human body hierarchical structure, we devise a graph pyramid module (GPM) by stacking three levels of graph structures from coarse granularity to fine granularity subsequently. At each level, GPM utilizes the self-attention mechanism to model the correlations between context nodes. Then, it adopts a top-down mechanism to progressively refine the hierarchical features through all the levels. GPM also enables efficient mutual learning. Specifically, the network weights of the first two levels are shared to exchange the learned coarse-granularity information across different datasets. By making use of the multi-granularity labels, Grapy-ML learns a more discriminative feature representation and achieves state-of-the-art performance, which is demonstrated by extensive experiments on the three popular benchmarks, e.g. CIHP dataset. The source code is publicly available at https://github.com/Charleshhy/Grapy-ML.
Tasks	Human Parsing, Semantic Segmentation
Published	2019-11-27
URL	https://arxiv.org/abs/1911.12053v1
PDF	https://arxiv.org/pdf/1911.12053v1.pdf
PWC	https://paperswithcode.com/paper/grapy-ml-graph-pyramid-mutual-learning-for
Repo	https://github.com/Charleshhy/Grapy-ML
Framework	pytorch

Computação Urbana da Teoria à Prática: Fundamentos, Aplicações e Desafios


Title	Computação Urbana da Teoria à Prática: Fundamentos, Aplicações e Desafios
Authors	Diego O. Rodrigues, Frances A. Santos, Geraldo P. Rocha Filho, Ademar T. Akabane, Raquel Cabral, Roger Immich, Wellington L. Junior, Felipe D. Cunha, Daniel L. Guidoni, Thiago H. Silva, Denis Rosário, Eduardo Cerqueira, Antonio A. F. Loureiro, Leandro A. Villas
Abstract	The growing of cities has resulted in innumerable technical and managerial challenges for public administrators such as energy consumption, pollution, urban mobility and even supervision of private and public spaces in an appropriate way. Urban Computing emerges as a promising paradigm to solve such challenges, through the extraction of knowledge, from a large amount of heterogeneous data existing in urban space. Moreover, Urban Computing correlates urban sensing, data management, and analysis to provide services that have the potential to improve the quality of life of the citizens of large urban centers. Consider this context, this chapter aims to present the fundamentals of Urban Computing and the steps necessary to develop an application in this area. To achieve this goal, the following questions will be investigated, namely: (i) What are the main research problems of Urban Computing?; (ii) What are the technological challenges for the implementation of services in Urban Computing?; (iii) What are the main methodologies used for the development of services in Urban Computing?; and (iv) What are the representative applications in this field?
Tasks
Published	2019-12-02
URL	https://arxiv.org/abs/1912.05662v1
PDF	https://arxiv.org/pdf/1912.05662v1.pdf
PWC	https://paperswithcode.com/paper/computacao-urbana-da-teoria-a-pratica
Repo	https://github.com/diegopso/hybrid-urban-routing-tutorial-sbrc
Framework	none

Giveme5W1H: A Universal System for Extracting Main Events from News Articles


Title	Giveme5W1H: A Universal System for Extracting Main Events from News Articles
Authors	Felix Hamborg, Corinna Breitinger, Bela Gipp
Abstract	Event extraction from news articles is a commonly required prerequisite for various tasks, such as article summarization, article clustering, and news aggregation. Due to the lack of universally applicable and publicly available methods tailored to news datasets, many researchers redundantly implement event extraction methods for their own projects. The journalistic 5W1H questions are capable of describing the main event of an article, i.e., by answering who did what, when, where, why, and how. We provide an in-depth description of an improved version of Giveme5W1H, a system that uses syntactic and domain-specific rules to automatically extract the relevant phrases from English news articles to provide answers to these 5W1H questions. Given the answers to these questions, the system determines an article’s main event. In an expert evaluation with three assessors and 120 articles, we determined an overall precision of p=0.73, and p=0.82 for answering the first four W questions, which alone can sufficiently summarize the main event reported on in a news article. We recently made our system publicly available, and it remains the only universal open-source 5W1H extractor capable of being applied to a wide range of use cases in news analysis.
Tasks
Published	2019-09-06
URL	https://arxiv.org/abs/1909.02766v1
PDF	https://arxiv.org/pdf/1909.02766v1.pdf
PWC	https://paperswithcode.com/paper/giveme5w1h-a-universal-system-for-extracting
Repo	https://github.com/fhamborg/Giveme5W1H
Framework	none

Adversarial point perturbations on 3D objects


Title	Adversarial point perturbations on 3D objects
Authors	Daniel Liu, Ronald Yu, Hao Su
Abstract	The importance of training robust neural network grows as 3D data is increasingly utilized in deep learning for vision tasks, like autonomous driving. We examine this problem from the perspective of the attacker, which is necessary in understanding how neural networks can be exploited, and thus defended. More specifically, we propose adversarial attacks based on solving different optimization problems, like minimizing the perceptibility of our generated adversarial examples, or maintaining a uniform density distribution of points across the adversarial object surfaces. Our four proposed algorithms for attacking 3D point cloud classification are all highly successful on existing neural networks, and we find that some of them are even effective against previously proposed point removal defenses.
Tasks	Autonomous Driving
Published	2019-08-16
URL	https://arxiv.org/abs/1908.06062v1
PDF	https://arxiv.org/pdf/1908.06062v1.pdf
PWC	https://paperswithcode.com/paper/adversarial-point-perturbations-on-3d-objects
Repo	https://github.com/Daniel-Liu-c0deb0t/Adversarial-point-perturbations-on-3D-objects
Framework	tf

StructureNet: Hierarchical Graph Networks for 3D Shape Generation


Title	StructureNet: Hierarchical Graph Networks for 3D Shape Generation
Authors	Kaichun Mo, Paul Guerrero, Li Yi, Hao Su, Peter Wonka, Niloy Mitra, Leonidas J. Guibas
Abstract	The ability to generate novel, diverse, and realistic 3D shapes along with associated part semantics and structure is central to many applications requiring high-quality 3D assets or large volumes of realistic training data. A key challenge towards this goal is how to accommodate diverse shape variations, including both continuous deformations of parts as well as structural or discrete alterations which add to, remove from, or modify the shape constituents and compositional structure. Such object structure can typically be organized into a hierarchy of constituent object parts and relationships, represented as a hierarchy of n-ary graphs. We introduce StructureNet, a hierarchical graph network which (i) can directly encode shapes represented as such n-ary graphs; (ii) can be robustly trained on large and complex shape families; and (iii) can be used to generate a great diversity of realistic structured shape geometries. Technically, we accomplish this by drawing inspiration from recent advances in graph neural networks to propose an order-invariant encoding of n-ary graphs, considering jointly both part geometry and inter-part relations during network training. We extensively evaluate the quality of the learned latent spaces for various shape families and show significant advantages over baseline and competing methods. The learned latent spaces enable several structure-aware geometry processing applications, including shape generation and interpolation, shape editing, or shape structure discovery directly from un-annotated images, point clouds, or partial scans.
Tasks	3D Shape Generation
Published	2019-08-01
URL	https://arxiv.org/abs/1908.00575v1
PDF	https://arxiv.org/pdf/1908.00575v1.pdf
PWC	https://paperswithcode.com/paper/structurenet-hierarchical-graph-networks-for
Repo	https://github.com/daerduoCarey/structurenet
Framework	pytorch

3DN: 3D Deformation Network


Title	3DN: 3D Deformation Network
Authors	Weiyue Wang, Duygu Ceylan, Radomir Mech, Ulrich Neumann
Abstract	Applications in virtual and augmented reality create a demand for rapid creation and easy access to large sets of 3D models. An effective way to address this demand is to edit or deform existing 3D models based on a reference, e.g., a 2D image which is very easy to acquire. Given such a source 3D model and a target which can be a 2D image, 3D model, or a point cloud acquired as a depth scan, we introduce 3DN, an end-to-end network that deforms the source model to resemble the target. Our method infers per-vertex offset displacements while keeping the mesh connectivity of the source model fixed. We present a training strategy which uses a novel differentiable operation, mesh sampling operator, to generalize our method across source and target models with varying mesh densities. Mesh sampling operator can be seamlessly integrated into the network to handle meshes with different topologies. Qualitative and quantitative results show that our method generates higher quality results compared to the state-of-the art learning-based methods for 3D shape generation. Code is available at github.com/laughtervv/3DN.
Tasks	3D Shape Generation
Published	2019-03-08
URL	http://arxiv.org/abs/1903.03322v1
PDF	http://arxiv.org/pdf/1903.03322v1.pdf
PWC	https://paperswithcode.com/paper/3dn-3d-deformation-network
Repo	https://github.com/laughtervv/3DN
Framework	tf

EmotionX-KU: BERT-Max based Contextual Emotion Classifier


Title	EmotionX-KU: BERT-Max based Contextual Emotion Classifier
Authors	Kisu Yang, Dongyub Lee, Taesun Whang, Seolhwa Lee, Heuiseok Lim
Abstract	We propose a contextual emotion classifier based on a transferable language model and dynamic max pooling, which predicts the emotion of each utterance in a dialogue. A representative emotion analysis task, EmotionX, requires to consider contextual information from colloquial dialogues and to deal with a class imbalance problem. To alleviate these problems, our model leverages the self-attention based transferable language model and the weighted cross entropy loss. Furthermore, we apply post-training and fine-tuning mechanisms to enhance the domain adaptability of our model and utilize several machine learning techniques to improve its performance. We conduct experiments on two emotion-labeled datasets named Friends and EmotionPush. As a result, our model outperforms the previous state-of-the-art model and also shows competitive performance in the EmotionX 2019 challenge. The code will be available in the Github page.
Tasks	Emotion Recognition, Language Modelling
Published	2019-06-27
URL	https://arxiv.org/abs/1906.11565v2
PDF	https://arxiv.org/pdf/1906.11565v2.pdf
PWC	https://paperswithcode.com/paper/emotionx-ku-bert-max-based-contextual-emotion
Repo	https://github.com/KisuYang/EmotionX-KU
Framework	pytorch

Wasserstein Weisfeiler-Lehman Graph Kernels


Title	Wasserstein Weisfeiler-Lehman Graph Kernels
Authors	Matteo Togninalli, Elisabetta Ghisu, Felipe Llinares-López, Bastian Rieck, Karsten Borgwardt
Abstract	Most graph kernels are an instance of the class of $\mathcal{R}$-Convolution kernels, which measure the similarity of objects by comparing their substructures. Despite their empirical success, most graph kernels use a naive aggregation of the final set of substructures, usually a sum or average, thereby potentially discarding valuable information about the distribution of individual components. Furthermore, only a limited instance of these approaches can be extended to continuously attributed graphs. We propose a novel method that relies on the Wasserstein distance between the node feature vector distributions of two graphs, which allows to find subtler differences in data sets by considering graphs as high-dimensional objects, rather than simple means. We further propose a Weisfeiler-Lehman inspired embedding scheme for graphs with continuous node attributes and weighted edges, enhance it with the computed Wasserstein distance, and thus improve the state-of-the-art prediction performance on several graph classification tasks.
Tasks	Graph Classification
Published	2019-06-04
URL	https://arxiv.org/abs/1906.01277v2
PDF	https://arxiv.org/pdf/1906.01277v2.pdf
PWC	https://paperswithcode.com/paper/wasserstein-weisfeiler-lehman-graph-kernels
Repo	https://github.com/BorgwardtLab/WWL
Framework	none

EM with Bias-Corrected Calibration is Hard-To-Beat at Label Shift Adaptation


Title	EM with Bias-Corrected Calibration is Hard-To-Beat at Label Shift Adaptation
Authors	Amr Alexandari, Anshul Kundaje, Avanti Shrikumar
Abstract	Label shift refers to the phenomenon where the prior class probability p(y) changes between the training and test distributions, while the conditional probability p(xy) stays fixed. Label shift arises in settings like medical diagnosis, where a classifier trained to predict disease given symptoms must be adapted to scenarios where the baseline prevalence of the disease is different. Given estimates of p(yx) from a predictive model, Saerens et al. (2002) proposed an efficient EM algorithm to correct for label shift that does not require model retraining. A limiting assumption of this algorithm is that p(yx) is calibrated, which is not true of modern neural networks. Recently, Black Box Shift Learning (BBSL) (Lipton et al., 2018) and Regularized Learning under Label Shifts (RLLS) (Azizzadenesheli et al., 2019) have emerged as state-of-the-art techniques to cope with label shift when a classifier does not output calibrated probabilities. However, both BBSL and RLLS require model retraining with importance weights, which poses challenges in practice (Byrd and Lipton, 2019), and neither has been benchmarked against EM. Here we show that by combining EM with a type of calibration we call bias-corrected calibration, we outperform both BBSL and RLLS across diverse datasets and distribution shifts. We further show that the EM objective is concave and bounded, and introduce a theoretically principled strategy for estimating source-domain priors that improves robustness to poor calibration. This work demonstrates that EM with appropriate calibration is a formidable and efficient baseline that future work in label shift adaptation should be compared against. Colab notebooks reproducing experiments are available at (anonymized link): https://github.com/blindauth/labelshiftexperiments
Tasks	Calibration, Diabetic Retinopathy Detection, Domain Adaptation, Image Classification, Medical Diagnosis
Published	2019-01-21
URL	https://arxiv.org/abs/1901.06852v4
PDF	https://arxiv.org/pdf/1901.06852v4.pdf
PWC	https://paperswithcode.com/paper/calibration-with-bias-corrected-temperature
Repo	https://github.com/kundajelab/abstention
Framework	none

Sionnx: Automatic Unit Test Generator for ONNX Conformance


Title	Sionnx: Automatic Unit Test Generator for ONNX Conformance
Authors	Xinli Cai, Peng Zhou, Shuhan Ding, Guoyang Chen, Weifeng Zhang
Abstract	Open Neural Network Exchange (ONNX) is an open format to represent AI models and is supported by many machine learning frameworks. While ONNX defines unified and portable computation operators across various frameworks, the conformance tests for those operators are insufficient, which makes it difficult to verify if an operator’s behavior in an ONNX backend implementation complies with the ONNX standard. In this paper, we present the first automatic unit test generator named Sionnx for verifying the compliance of ONNX implementation. First, we propose a compact yet complete set of rules to describe the operator’s attributes and the properties of its operands. Second, we design an Operator Specification Language (OSL) to provide a high-level description for the operator’s syntax. Finally, through this easy-to-use specification language, we are able to build a full testing specification which leverages LLVM TableGen to automatically generate unit tests for ONNX operators with much large coverage. Sionnx is lightweight and flexible to support cross-framework verification. The Sionnx framework is open-sourced in the github repository (https://github.com/alibaba/Sionnx).
Tasks
Published	2019-06-12
URL	https://arxiv.org/abs/1906.05676v1
PDF	https://arxiv.org/pdf/1906.05676v1.pdf
PWC	https://paperswithcode.com/paper/sionnx-automatic-unit-test-generator-for-onnx
Repo	https://github.com/alibaba/Sionnx
Framework	none

MultiResUNet : Rethinking the U-Net Architecture for Multimodal Biomedical Image Segmentation


Title	MultiResUNet : Rethinking the U-Net Architecture for Multimodal Biomedical Image Segmentation
Authors	Nabil Ibtehaz, M. Sohel Rahman
Abstract	In recent years Deep Learning has brought about a breakthrough in Medical Image Segmentation. U-Net is the most prominent deep network in this regard, which has been the most popular architecture in the medical imaging community. Despite outstanding overall performance in segmenting multimodal medical images, from extensive experimentations on challenging datasets, we found out that the classical U-Net architecture seems to be lacking in certain aspects. Therefore, we propose some modifications to improve upon the already state-of-the-art U-Net model. Hence, following the modifications we develop a novel architecture MultiResUNet as the potential successor to the successful U-Net architecture. We have compared our proposed architecture MultiResUNet with the classical U-Net on a vast repertoire of multimodal medical images. Albeit slight improvements in the cases of ideal images, a remarkable gain in performance has been attained for challenging images. We have evaluated our model on five different datasets, each with their own unique challenges, and have obtained a relative improvement in performance of 10.15%, 5.07%, 2.63%, 1.41%, and 0.62% respectively.
Tasks	Medical Image Segmentation, Semantic Segmentation
Published	2019-02-11
URL	http://arxiv.org/abs/1902.04049v1
PDF	http://arxiv.org/pdf/1902.04049v1.pdf
PWC	https://paperswithcode.com/paper/multiresunet-rethinking-the-u-net
Repo	https://github.com/nibtehaz/MultiResUNet
Framework	tf

Guided Visual Exploration of Relations in Data Sets


Title	Guided Visual Exploration of Relations in Data Sets
Authors	Kai Puolamäki, Emilia Oikarinen, Andreas Henelius
Abstract	Efficient explorative data analysis systems must take into account both what a user knows and wants to know. This paper proposes a principled framework for interactive visual exploration of relations in data, through views most informative given the user’s current knowledge and objectives. The user can input pre-existing knowledge of relations in the data and also formulate specific exploration interests, then taken into account in the exploration. The idea is to steer the exploration process towards the interests of the user, instead of showing uninteresting or already known relations. The user’s knowledge is modelled by a distribution over data sets parametrised by subsets of rows and columns of data, called tile constraints. We provide a computationally efficient implementation of this concept based on constrained randomisation. Furthermore, we describe a novel dimensionality reduction method for finding the views most informative to the user, which at the limit of no background knowledge and with generic objectives reduces to PCA. We show that the method is suitable for interactive use and robust to noise, outperforms standard projection pursuit visualisation methods, and gives understandable and useful results in analysis of real-world data. We have released an open-source implementation of the framework.
Tasks	Dimensionality Reduction
Published	2019-05-07
URL	https://arxiv.org/abs/1905.02515v1
PDF	https://arxiv.org/pdf/1905.02515v1.pdf
PWC	https://paperswithcode.com/paper/guided-visual-exploration-of-relations-in
Repo	https://github.com/edahelsinki/corand
Framework	none

When Causal Intervention Meets Adversarial Examples and Image Masking for Deep Neural Networks


Title	When Causal Intervention Meets Adversarial Examples and Image Masking for Deep Neural Networks
Authors	Chao-Han Huck Yang, Yi-Chieh Liu, Pin-Yu Chen, Xiaoli Ma, Yi-Chang James Tsai
Abstract	Discovering and exploiting the causality in deep neural networks (DNNs) are crucial challenges for understanding and reasoning causal effects (CE) on an explainable visual model. “Intervention” has been widely used for recognizing a causal relation ontologically. In this paper, we propose a causal inference framework for visual reasoning via do-calculus. To study the intervention effects on pixel-level features for causal reasoning, we introduce pixel-wise masking and adversarial perturbation. In our framework, CE is calculated using features in a latent space and perturbed prediction from a DNN-based model. We further provide the first look into the characteristics of discovered CE of adversarially perturbed images generated by gradient-based methods \footnote{~~https://github.com/jjaacckkyy63/Causal-Intervention-AE-wAdvImg}. Experimental results show that CE is a competitive and robust index for understanding DNNs when compared with conventional methods such as class-activation mappings (CAMs) on the Chest X-Ray-14 dataset for human-interpretable feature(s) (e.g., symptom) reasoning. Moreover, CE holds promises for detecting adversarial examples as it possesses distinct characteristics in the presence of adversarial perturbations.
Tasks	Causal Inference, Visual Reasoning
Published	2019-02-09
URL	https://arxiv.org/abs/1902.03380v3
PDF	https://arxiv.org/pdf/1902.03380v3.pdf
PWC	https://paperswithcode.com/paper/when-causal-intervention-meets-image-masking
Repo	https://github.com/jjaacckkyy63/Causal-Intervention-AE-wAdvImg
Framework	none

Keep It Simple: Graph Autoencoders Without Graph Convolutional Networks


Title	Keep It Simple: Graph Autoencoders Without Graph Convolutional Networks
Authors	Guillaume Salha, Romain Hennequin, Michalis Vazirgiannis
Abstract	Graph autoencoders (AE) and variational autoencoders (VAE) recently emerged as powerful node embedding methods, with promising performances on challenging tasks such as link prediction and node clustering. Graph AE, VAE and most of their extensions rely on graph convolutional networks (GCN) to learn vector space representations of nodes. In this paper, we propose to replace the GCN encoder by a simple linear model w.r.t. the adjacency matrix of the graph. For the two aforementioned tasks, we empirically show that this approach consistently reaches competitive performances w.r.t. GCN-based models for numerous real-world graphs, including the widely used Cora, Citeseer and Pubmed citation networks that became the de facto benchmark datasets for evaluating graph AE and VAE. This result questions the relevance of repeatedly using these three datasets to compare complex graph AE and VAE models. It also emphasizes the effectiveness of simple node encoding schemes for many real-world applications.
Tasks	Link Prediction
Published	2019-10-02
URL	https://arxiv.org/abs/1910.00942v1
PDF	https://arxiv.org/pdf/1910.00942v1.pdf
PWC	https://paperswithcode.com/paper/keep-it-simple-graph-autoencoders-without
Repo	https://github.com/deezer/linear_graph_autoencoders
Framework	tf

Provably Robust Boosted Decision Stumps and Trees against Adversarial Attacks


Title	Provably Robust Boosted Decision Stumps and Trees against Adversarial Attacks
Authors	Maksym Andriushchenko, Matthias Hein
Abstract	The problem of adversarial robustness has been studied extensively for neural networks. However, for boosted decision trees and decision stumps there are almost no results, even though they are widely used in practice (e.g. XGBoost) due to their accuracy, interpretability, and efficiency. We show in this paper that for boosted decision stumps the \textit{exact} min-max robust loss and test error for an $l_\infty$-attack can be computed in $O(T\log T)$ time per input, where $T$ is the number of decision stumps and the optimal update step of the ensemble can be done in $O(n^2,T\log T)$, where $n$ is the number of data points. For boosted trees we show how to efficiently calculate and optimize an upper bound on the robust loss, which leads to state-of-the-art robust test error for boosted trees on MNIST (12.5% for $\epsilon_\infty=0.3$), FMNIST (23.2% for $\epsilon_\infty=0.1$), and CIFAR-10 (74.7% for $\epsilon_\infty=8/255$). Moreover, the robust test error rates we achieve are competitive to the ones of provably robust convolutional networks. The code of all our experiments is available at http://github.com/max-andr/provably-robust-boosting
Tasks
Published	2019-06-08
URL	https://arxiv.org/abs/1906.03526v2
PDF	https://arxiv.org/pdf/1906.03526v2.pdf
PWC	https://paperswithcode.com/paper/provably-robust-boosted-decision-stumps-and
Repo	https://github.com/max-andr/provably-robust-boosting
Framework	none