October 20, 2019

3075 words 15 mins read

Paper Group AWR 253

SAVOIAS: A Diverse, Multi-Category Visual Complexity Dataset. Texygen: A Benchmarking Platform for Text Generation Models. 3D Consistent & Robust Segmentation of Cardiac Images by Deep Learning with Spatial Propagation. Learning to Count Objects in Natural Images for Visual Question Answering. PaMpeR: Proof Method Recommendation System for Isabelle …

SAVOIAS: A Diverse, Multi-Category Visual Complexity Dataset


Title	SAVOIAS: A Diverse, Multi-Category Visual Complexity Dataset
Authors	Elham Saraee, Mona Jalal, Margrit Betke
Abstract	Visual complexity identifies the level of intricacy and details in an image or the level of difficulty to describe the image. It is an important concept in a variety of areas such as cognitive psychology, computer vision and visualization, and advertisement. Yet, efforts to create large, downloadable image datasets with diverse content and unbiased groundtruthing are lacking. In this work, we introduce Savoias, a visual complexity dataset that compromises of more than 1,400 images from seven image categories relevant to the above research areas, namely Scenes, Advertisements, Visualization and infographics, Objects, Interior design, Art, and Suprematism. The images in each category portray diverse characteristics including various low-level and high-level features, objects, backgrounds, textures and patterns, text, and graphics. The ground truth for Savoias is obtained by crowdsourcing more than 37,000 pairwise comparisons of images using the forced-choice methodology and with more than 1,600 contributors. The resulting relative scores are then converted to absolute visual complexity scores using the Bradley-Terry method and matrix completion. When applying five state-of-the-art algorithms to analyze the visual complexity of the images in the Savoias dataset, we found that the scores obtained from these baseline tools only correlate well with crowdsourced labels for abstract patterns in the Suprematism category (Pearson correlation r=0.84). For the other categories, in particular, the objects and advertisement categories, low correlation coefficients were revealed (r=0.3 and 0.56, respectively). These findings suggest that (1) state-of-the-art approaches are mostly insufficient and (2) Savoias enables category-specific method development, which is likely to improve the impact of visual complexity analysis on specific application areas, including computer vision.
Tasks	Matrix Completion
Published	2018-10-03
URL	http://arxiv.org/abs/1810.01771v1
PDF	http://arxiv.org/pdf/1810.01771v1.pdf
PWC	https://paperswithcode.com/paper/savoias-a-diverse-multi-category-visual
Repo	https://github.com/esaraee/Savoias-Dataset
Framework	none

Texygen: A Benchmarking Platform for Text Generation Models


Title	Texygen: A Benchmarking Platform for Text Generation Models
Authors	Yaoming Zhu, Sidi Lu, Lei Zheng, Jiaxian Guo, Weinan Zhang, Jun Wang, Yong Yu
Abstract	We introduce Texygen, a benchmarking platform to support research on open-domain text generation models. Texygen has not only implemented a majority of text generation models, but also covered a set of metrics that evaluate the diversity, the quality and the consistency of the generated texts. The Texygen platform could help standardize the research on text generation and facilitate the sharing of fine-tuned open-source implementations among researchers for their work. As a consequence, this would help in improving the reproductivity and reliability of future research work in text generation.
Tasks	Text Generation
Published	2018-02-06
URL	http://arxiv.org/abs/1802.01886v1
PDF	http://arxiv.org/pdf/1802.01886v1.pdf
PWC	https://paperswithcode.com/paper/texygen-a-benchmarking-platform-for-text
Repo	https://github.com/geek-ai/Texygen
Framework	tf

3D Consistent & Robust Segmentation of Cardiac Images by Deep Learning with Spatial Propagation


Title	3D Consistent & Robust Segmentation of Cardiac Images by Deep Learning with Spatial Propagation
Authors	Qiao Zheng, Hervé Delingette, Nicolas Duchateau, Nicholas Ayache
Abstract	We propose a method based on deep learning to perform cardiac segmentation on short axis MRI image stacks iteratively from the top slice (around the base) to the bottom slice (around the apex). At each iteration, a novel variant of U-net is applied to propagate the segmentation of a slice to the adjacent slice below it. In other words, the prediction of a segmentation of a slice is dependent upon the already existing segmentation of an adjacent slice. 3D-consistency is hence explicitly enforced. The method is trained on a large database of 3078 cases from UK Biobank. It is then tested on 756 different cases from UK Biobank and three other state-of-the-art cohorts (ACDC with 100 cases, Sunnybrook with 30 cases, RVSC with 16 cases). Results comparable or even better than the state-of-the-art in terms of distance measures are achieved. They also emphasize the assets of our method, namely enhanced spatial consistency (currently neither considered nor achieved by the state-of-the-art), and the generalization ability to unseen cases even from other databases.
Tasks	Cardiac Segmentation
Published	2018-04-25
URL	http://arxiv.org/abs/1804.09400v1
PDF	http://arxiv.org/pdf/1804.09400v1.pdf
PWC	https://paperswithcode.com/paper/3d-consistent-robust-segmentation-of-cardiac
Repo	https://github.com/julien-zheng/CardiacSegmentationPropagation
Framework	tf

Learning to Count Objects in Natural Images for Visual Question Answering


Title	Learning to Count Objects in Natural Images for Visual Question Answering
Authors	Yan Zhang, Jonathon Hare, Adam Prügel-Bennett
Abstract	Visual Question Answering (VQA) models have struggled with counting objects in natural images so far. We identify a fundamental problem due to soft attention in these models as a cause. To circumvent this problem, we propose a neural network component that allows robust counting from object proposals. Experiments on a toy task show the effectiveness of this component and we obtain state-of-the-art accuracy on the number category of the VQA v2 dataset without negatively affecting other categories, even outperforming ensemble models with our single model. On a difficult balanced pair metric, the component gives a substantial improvement in counting over a strong baseline by 6.6%.
Tasks	Visual Question Answering
Published	2018-02-15
URL	http://arxiv.org/abs/1802.05766v1
PDF	http://arxiv.org/pdf/1802.05766v1.pdf
PWC	https://paperswithcode.com/paper/learning-to-count-objects-in-natural-images
Repo	https://github.com/Cyanogenoid/vqa-counting
Framework	pytorch

PaMpeR: Proof Method Recommendation System for Isabelle/HOL


Title	PaMpeR: Proof Method Recommendation System for Isabelle/HOL
Authors	Yutaka Nagashima, Yilun He
Abstract	Deciding which sub-tool to use for a given proof state requires expertise specific to each ITP. To mitigate this problem, we present PaMpeR, a Proof Method Recommendation system for Isabelle/HOL. Given a proof state, PaMpeR recommends proof methods to discharge the proof goal and provides qualitative explanations as to why it suggests these methods. PaMpeR generates these recommendations based on existing hand-written proof corpora, thus transferring experienced users’ expertise to new users. Our evaluation shows that PaMpeR correctly predicts experienced users’ proof methods invocation especially when it comes to special purpose proof methods.
Tasks
Published	2018-06-19
URL	http://arxiv.org/abs/1806.07239v1
PDF	http://arxiv.org/pdf/1806.07239v1.pdf
PWC	https://paperswithcode.com/paper/pamper-proof-method-recommendation-system-for
Repo	https://github.com/data61/PSL
Framework	none

Optimal ridge penalty for real-world high-dimensional data can be zero or negative due to the implicit ridge regularization


Title	Optimal ridge penalty for real-world high-dimensional data can be zero or negative due to the implicit ridge regularization
Authors	Dmitry Kobak, Jonathan Lomond, Benoit Sanchez
Abstract	A conventional wisdom in statistical learning is that large models require strong regularization to prevent overfitting. Here we show that this rule can be violated by linear regression in the underdetermined $n\ll p$ situation under realistic conditions. Using simulations and real-life high-dimensional data sets, we demonstrate that an explicit positive ridge penalty can fail to provide any improvement over the minimum-norm least squares estimator. Moreover, the optimal value of ridge penalty in this situation can be negative. This happens when the high-variance directions in the predictor space can predict the response variable, which is often the case in the real-world high-dimensional data. In this regime, the low-variance directions provide an implicit ridge regularization and can make any further positive ridge penalty detrimental. We prove that augmenting any linear model with small random covariates and using minimum-norm estimator is asymptotically equivalent to adding the ridge penalty.
Tasks
Published	2018-05-28
URL	https://arxiv.org/abs/1805.10939v3
PDF	https://arxiv.org/pdf/1805.10939v3.pdf
PWC	https://paperswithcode.com/paper/implicit-ridge-regularization-provided-by-the
Repo	https://github.com/dkobak/high-dim-ridge
Framework	none

Deep Learning of Vortex Induced Vibrations


Title	Deep Learning of Vortex Induced Vibrations
Authors	Maziar Raissi, Zhicheng Wang, Michael S. Triantafyllou, George Em Karniadakis
Abstract	Vortex induced vibrations of bluff bodies occur when the vortex shedding frequency is close to the natural frequency of the structure. Of interest is the prediction of the lift and drag forces on the structure given some limited and scattered information on the velocity field. This is an inverse problem that is not straightforward to solve using standard computational fluid dynamics (CFD) methods, especially since no information is provided for the pressure. An even greater challenge is to infer the lift and drag forces given some dye or smoke visualizations of the flow field. Here we employ deep neural networks that are extended to encode the incompressible Navier-Stokes equations coupled with the structure’s dynamic motion equation. In the first case, given scattered data in space-time on the velocity field and the structure’s motion, we use four coupled deep neural networks to infer very accurately the structural parameters, the entire time-dependent pressure field (with no prior training data), and reconstruct the velocity vector field and the structure’s dynamic motion. In the second case, given scattered data in space-time on a concentration field only, we use five coupled deep neural networks to infer very accurately the vector velocity field and all other quantities of interest as before. This new paradigm of inference in fluid mechanics for coupled multi-physics problems enables velocity and pressure quantification from flow snapshots in small subdomains and can be exploited for flow control applications and also for system identification.
Tasks
Published	2018-08-26
URL	http://arxiv.org/abs/1808.08952v1
PDF	http://arxiv.org/pdf/1808.08952v1.pdf
PWC	https://paperswithcode.com/paper/deep-learning-of-vortex-induced-vibrations
Repo	https://github.com/maziarraissi/DeepVIV
Framework	tf

Learning Transferable Adversarial Examples via Ghost Networks


Title	Learning Transferable Adversarial Examples via Ghost Networks
Authors	Yingwei Li, Song Bai, Yuyin Zhou, Cihang Xie, Zhishuai Zhang, Alan Yuille
Abstract	Recent development of adversarial attacks has proven that ensemble-based methods outperform traditional, non-ensemble ones in black-box attack. However, as it is computationally prohibitive to acquire a family of diverse models, these methods achieve inferior performance constrained by the limited number of models to be ensembled. In this paper, we propose Ghost Networks to improve the transferability of adversarial examples. The critical principle of ghost networks is to apply feature-level perturbations to an existing model to potentially create a huge set of diverse models. After that, models are subsequently fused by longitudinal ensemble. Extensive experimental results suggest that the number of networks is essential for improving the transferability of adversarial examples, but it is less necessary to independently train different networks and ensemble them in an intensive aggregation way. Instead, our work can be used as a computationally cheap and easily applied plug-in to improve adversarial approaches both in single-model and multi-model attack, compatible with residual and non-residual networks. By reproducing the NeurIPS 2017 adversarial competition, our method outperforms the No.1 attack submission by a large margin, demonstrating its effectiveness and efficiency. Code is available at https://github.com/LiYingwei/ghost-network.
Tasks	Adversarial Attack
Published	2018-12-09
URL	https://arxiv.org/abs/1812.03413v3
PDF	https://arxiv.org/pdf/1812.03413v3.pdf
PWC	https://paperswithcode.com/paper/learning-transferable-adversarial-examples
Repo	https://github.com/LiYingwei/ghost-network
Framework	tf

M-ADDA: Unsupervised Domain Adaptation with Deep Metric Learning


Title	M-ADDA: Unsupervised Domain Adaptation with Deep Metric Learning
Authors	Issam Laradji, Reza Babanezhad
Abstract	Unsupervised domain adaptation techniques have been successful for a wide range of problems where supervised labels are limited. The task is to classify an unlabeled `target' dataset by leveraging a labeled` source’ dataset that comes from a slightly similar distribution. We propose metric-based adversarial discriminative domain adaptation (M-ADDA) which performs two main steps. First, it uses a metric learning approach to train the source model on the source dataset by optimizing the triplet loss function. This results in clusters where embeddings of the same label are close to each other and those with different labels are far from one another. Next, it uses the adversarial approach (as that used in ADDA \cite{2017arXiv170205464T}) to make the extracted features from the source and target datasets indistinguishable. Simultaneously, we optimize a novel loss function that encourages the target dataset’s embeddings to form clusters. While ADDA and M-ADDA use similar architectures, we show that M-ADDA performs significantly better on the digits adaptation datasets of MNIST and USPS. This suggests that using metric-learning for domain adaptation can lead to large improvements in classification accuracy for the domain adaptation task. The code is available at \url{https://github.com/IssamLaradji/M-ADDA}.
Tasks	Domain Adaptation, Metric Learning, Unsupervised Domain Adaptation
Published	2018-07-06
URL	http://arxiv.org/abs/1807.02552v1
PDF	http://arxiv.org/pdf/1807.02552v1.pdf
PWC	https://paperswithcode.com/paper/m-adda-unsupervised-domain-adaptation-with
Repo	https://github.com/IssamLaradji/M-ADDA
Framework	pytorch

End-Task Oriented Textual Entailment via Deep Explorations of Inter-Sentence Interactions


Title	End-Task Oriented Textual Entailment via Deep Explorations of Inter-Sentence Interactions
Authors	Wenpeng Yin, Hinrich Schütze, Dan Roth
Abstract	This work deals with SciTail, a natural entailment challenge derived from a multi-choice question answering problem. The premises and hypotheses in SciTail were generated with no awareness of each other, and did not specifically aim at the entailment task. This makes it more challenging than other entailment data sets and more directly useful to the end-task – question answering. We propose DEISTE (deep explorations of inter-sentence interactions for textual entailment) for this entailment task. Given word-to-word interactions between the premise-hypothesis pair ($P$, $H$), DEISTE consists of: (i) a parameter-dynamic convolution to make important words in $P$ and $H$ play a dominant role in learnt representations; and (ii) a position-aware attentive convolution to encode the representation and position information of the aligned word pairs. Experiments show that DEISTE gets $\approx$5% improvement over prior state of the art and that the pretrained DEISTE on SciTail generalizes well on RTE-5.
Tasks	Natural Language Inference, Question Answering
Published	2018-04-24
URL	http://arxiv.org/abs/1804.08813v3
PDF	http://arxiv.org/pdf/1804.08813v3.pdf
PWC	https://paperswithcode.com/paper/end-task-oriented-textual-entailment-via-deep
Repo	https://github.com/yinwenpeng/SciTail
Framework	none

Incorporating Chinese Characters of Words for Lexical Sememe Prediction


Title	Incorporating Chinese Characters of Words for Lexical Sememe Prediction
Authors	Huiming Jin, Hao Zhu, Zhiyuan Liu, Ruobing Xie, Maosong Sun, Fen Lin, Leyu Lin
Abstract	Sememes are minimum semantic units of concepts in human languages, such that each word sense is composed of one or multiple sememes. Words are usually manually annotated with their sememes by linguists, and form linguistic common-sense knowledge bases widely used in various NLP tasks. Recently, the lexical sememe prediction task has been introduced. It consists of automatically recommending sememes for words, which is expected to improve annotation efficiency and consistency. However, existing methods of lexical sememe prediction typically rely on the external context of words to represent the meaning, which usually fails to deal with low-frequency and out-of-vocabulary words. To address this issue for Chinese, we propose a novel framework to take advantage of both internal character information and external context information of words. We experiment on HowNet, a Chinese sememe knowledge base, and demonstrate that our framework outperforms state-of-the-art baselines by a large margin, and maintains a robust performance even for low-frequency words.
Tasks	Common Sense Reasoning
Published	2018-06-17
URL	http://arxiv.org/abs/1806.06349v1
PDF	http://arxiv.org/pdf/1806.06349v1.pdf
PWC	https://paperswithcode.com/paper/incorporating-chinese-characters-of-words-for
Repo	https://github.com/thunlp/Character-enhanced-Sememe-Prediction
Framework	none

Automatic Detection of Neurons in NeuN-stained Histological Images of Human Brain


Title	Automatic Detection of Neurons in NeuN-stained Histological Images of Human Brain
Authors	Andrija Štajduhar, Domagoj Džaja, Miloš Judaš, Sven Lončarić
Abstract	In this paper, we present a novel use of an anisotropic diffusion model for automatic detection of neurons in histological sections of the adult human brain cortex. We use a partial differential equation model to process high resolution images to acquire locations of neuronal bodies. We also present a novel approach in model training and evaluation that considers variability among the human experts, addressing the issue of existence and correctness of the golden standard for neuron and cell counting, used in most of relevant papers. Our method, trained on dataset manually labeled by three experts, has correctly distinguished over 95% of neuron bodies in test data, doing so in time much shorter than other comparable methods.
Tasks
Published	2018-06-01
URL	http://arxiv.org/abs/1806.00292v1
PDF	http://arxiv.org/pdf/1806.00292v1.pdf
PWC	https://paperswithcode.com/paper/automatic-detection-of-neurons-in-neun
Repo	https://github.com/astajd/neurons
Framework	none

Fast Neural Architecture Construction using EnvelopeNets


Title	Fast Neural Architecture Construction using EnvelopeNets
Authors	Purushotham Kamath, Abhishek Singh, Debo Dutta
Abstract	Fast Neural Architecture Construction (NAC) is a method to construct deep network architectures by pruning and expansion of a base network. In recent years, several automated search methods for neural network architectures have been proposed using methods such as evolutionary algorithms and reinforcement learning. These methods use a single scalar objective function (usually accuracy) that is evaluated after a full training and evaluation cycle. In contrast NAC directly compares the utility of different filters using statistics derived from filter featuremaps reach a state where the utility of different filters within a network can be compared and hence can be used to construct networks. The training epochs needed for filters within a network to reach this state is much less than the training epochs needed for the accuracy of a network to stabilize. NAC exploits this finding to construct convolutional neural nets (CNNs) with close to state of the art accuracy, in < 1 GPU day, faster than most of the current neural architecture search methods. The constructed networks show close to state of the art performance on the image classification problem on well known datasets (CIFAR-10, ImageNet) and consistently show better performance than hand constructed and randomly generated networks of the same depth, operators and approximately the same number of parameters.
Tasks	Image Classification, Neural Architecture Search
Published	2018-03-18
URL	http://arxiv.org/abs/1803.06744v3
PDF	http://arxiv.org/pdf/1803.06744v3.pdf
PWC	https://paperswithcode.com/paper/fast-neural-architecture-construction-using
Repo	https://github.com/CiscoAI/amla
Framework	tf

Personalized Language Model for Query Auto-Completion


Title	Personalized Language Model for Query Auto-Completion
Authors	Aaron Jaech, Mari Ostendorf
Abstract	Query auto-completion is a search engine feature whereby the system suggests completed queries as the user types. Recently, the use of a recurrent neural network language model was suggested as a method of generating query completions. We show how an adaptable language model can be used to generate personalized completions and how the model can use online updating to make predictions for users not seen during training. The personalized predictions are significantly better than a baseline that uses no user information.
Tasks	Language Modelling
Published	2018-04-25
URL	http://arxiv.org/abs/1804.09661v1
PDF	http://arxiv.org/pdf/1804.09661v1.pdf
PWC	https://paperswithcode.com/paper/personalized-language-model-for-query-auto
Repo	https://github.com/ssharpe42/VNLQAC
Framework	tf

Improving Unsupervised Defect Segmentation by Applying Structural Similarity to Autoencoders


Title	Improving Unsupervised Defect Segmentation by Applying Structural Similarity to Autoencoders
Authors	Paul Bergmann, Sindy Löwe, Michael Fauser, David Sattlegger, Carsten Steger
Abstract	Convolutional autoencoders have emerged as popular methods for unsupervised defect segmentation on image data. Most commonly, this task is performed by thresholding a pixel-wise reconstruction error based on an $\ell^p$ distance. This procedure, however, leads to large residuals whenever the reconstruction encompasses slight localization inaccuracies around edges. It also fails to reveal defective regions that have been visually altered when intensity values stay roughly consistent. We show that these problems prevent these approaches from being applied to complex real-world scenarios and that it cannot be easily avoided by employing more elaborate architectures such as variational or feature matching autoencoders. We propose to use a perceptual loss function based on structural similarity which examines inter-dependencies between local image regions, taking into account luminance, contrast and structural information, instead of simply comparing single pixel values. It achieves significant performance gains on a challenging real-world dataset of nanofibrous materials and a novel dataset of two woven fabrics over the state of the art approaches for unsupervised defect segmentation that use pixel-wise reconstruction error metrics.
Tasks
Published	2018-07-05
URL	http://arxiv.org/abs/1807.02011v3
PDF	http://arxiv.org/pdf/1807.02011v3.pdf
PWC	https://paperswithcode.com/paper/improving-unsupervised-defect-segmentation-by
Repo	https://github.com/daxiaHuang/Unsupervised_Defect_Segmentation
Framework	tf