February 1, 2020

2983 words 15 mins read

Paper Group AWR 363

Spherical CNNs on Unstructured Grids. Bootstrapping the Expressivity with Model-based Planning. GAN-enhanced Conditional Echocardiogram Generation. Topic-aware chatbot using Recurrent Neural Networks and Nonnegative Matrix Factorization. Evaluation and Improvement of Chatbot Text Classification Data Quality Using Plausible Negative Examples. Guidin …

Spherical CNNs on Unstructured Grids


Title	Spherical CNNs on Unstructured Grids
Authors	Chiyu “Max” Jiang, Jingwei Huang, Karthik Kashinath, Prabhat, Philip Marcus, Matthias Niessner
Abstract	We present an efficient convolution kernel for Convolutional Neural Networks (CNNs) on unstructured grids using parameterized differential operators while focusing on spherical signals such as panorama images or planetary signals. To this end, we replace conventional convolution kernels with linear combinations of differential operators that are weighted by learnable parameters. Differential operators can be efficiently estimated on unstructured grids using one-ring neighbors, and learnable parameters can be optimized through standard back-propagation. As a result, we obtain extremely efficient neural networks that match or outperform state-of-the-art network architectures in terms of performance but with a significantly lower number of network parameters. We evaluate our algorithm in an extensive series of experiments on a variety of computer vision and climate science tasks, including shape classification, climate pattern segmentation, and omnidirectional image semantic segmentation. Overall, we present (1) a novel CNN approach on unstructured grids using parameterized differential operators for spherical signals, and (2) we show that our unique kernel parameterization allows our model to achieve the same or higher accuracy with significantly fewer network parameters.
Tasks	Semantic Segmentation
Published	2019-01-07
URL	http://arxiv.org/abs/1901.02039v1
PDF	http://arxiv.org/pdf/1901.02039v1.pdf
PWC	https://paperswithcode.com/paper/spherical-cnns-on-unstructured-grids
Repo	https://github.com/maxjiang93/ugscnn
Framework	pytorch

Bootstrapping the Expressivity with Model-based Planning


Title	Bootstrapping the Expressivity with Model-based Planning
Authors	Kefan Dong, Yuping Luo, Tengyu Ma
Abstract	We compare the model-free reinforcement learning with the model-based approaches through the lens of the expressive power of neural networks for policies, $Q$-functions, and dynamics. We show, theoretically and empirically, that even for one-dimensional continuous state space, there are many MDPs whose optimal $Q$-functions and policies are much more complex than the dynamics. We hypothesize many real-world MDPs also have a similar property. For these MDPs, model-based planning is a favorable algorithm, because the resulting policies can approximate the optimal policy significantly better than a neural network parameterization can, and model-free or model-based policy optimization rely on policy parameterization. Motivated by the theory, we apply a simple multi-step model-based bootstrapping planner (BOOTS) to bootstrap a weak $Q$-function into a stronger policy. Empirical results show that applying BOOTS on top of model-based or model-free policy optimization algorithms at the test time improves the performance on MuJoCo benchmark tasks.
Tasks
Published	2019-10-14
URL	https://arxiv.org/abs/1910.05927v1
PDF	https://arxiv.org/pdf/1910.05927v1.pdf
PWC	https://paperswithcode.com/paper/bootstrapping-the-expressivity-with-model
Repo	https://github.com/roosephu/boots
Framework	tf

GAN-enhanced Conditional Echocardiogram Generation


Title	GAN-enhanced Conditional Echocardiogram Generation
Authors	Amir H. Abdi, Teresa Tsang, Purang Abolmaesumi
Abstract	Echocardiography (echo) is a common means of evaluating cardiac conditions. Due to the label scarcity, semi-supervised paradigms in automated echo analysis are getting traction. One of the most sought-after problems in echo is the segmentation of cardiac structures (e.g. chambers). Accordingly, we propose an echocardiogram generation approach using generative adversarial networks with a conditional patch-based discriminator. In this work, we validate the feasibility of GAN-enhanced echo generation with different conditions (segmentation masks), namely, the left ventricle, ventricular myocardium, and atrium. Results show that the proposed adversarial algorithm can generate high-quality echo frames whose cardiac structures match the given segmentation masks. This method is expected to facilitate the training of other machine learning models in a semi-supervised fashion as suggested in similar researches.
Tasks
Published	2019-11-05
URL	https://arxiv.org/abs/1911.02121v2
PDF	https://arxiv.org/pdf/1911.02121v2.pdf
PWC	https://paperswithcode.com/paper/gan-enhanced-conditional-echocardiogram
Repo	https://github.com/amir-abdi/echo-generation
Framework	tf

Topic-aware chatbot using Recurrent Neural Networks and Nonnegative Matrix Factorization


Title	Topic-aware chatbot using Recurrent Neural Networks and Nonnegative Matrix Factorization
Authors	Yuchen Guo, Nicholas Hanoian, Zhexiao Lin, Nicholas Liskij, Hanbaek Lyu, Deanna Needell, Jiahao Qu, Henry Sojico, Yuliang Wang, Zhe Xiong, Zhenhong Zou
Abstract	We propose a novel model for a topic-aware chatbot by combining the traditional Recurrent Neural Network (RNN) encoder-decoder model with a topic attention layer based on Nonnegative Matrix Factorization (NMF). After learning topic vectors from an auxiliary text corpus via NMF, the decoder is trained so that it is more likely to sample response words from the most correlated topic vectors. One of the main advantages in our architecture is that the user can easily switch the NMF-learned topic vectors so that the chatbot obtains desired topic-awareness. We demonstrate our model by training on a single conversational data set which is then augmented with topic matrices learned from different auxiliary data sets. We show that our topic-aware chatbot not only outperforms the non-topic counterpart, but also that each topic-aware model qualitatively and contextually gives the most relevant answer depending on the topic of question.
Tasks	Chatbot
Published	2019-12-01
URL	https://arxiv.org/abs/1912.00315v2
PDF	https://arxiv.org/pdf/1912.00315v2.pdf
PWC	https://paperswithcode.com/paper/topic-aware-chatbot-using-recurrent-neural
Repo	https://github.com/HanbaekLyu/RNN_NMF_chatbot
Framework	pytorch

Evaluation and Improvement of Chatbot Text Classification Data Quality Using Plausible Negative Examples


Title	Evaluation and Improvement of Chatbot Text Classification Data Quality Using Plausible Negative Examples
Authors	Kit Kuksenok, Andriy Martyniv
Abstract	We describe and validate a metric for estimating multi-class classifier performance based on cross-validation and adapted for improvement of small, unbalanced natural-language datasets used in chatbot design. Our experiences draw upon building recruitment chatbots that mediate communication between job-seekers and recruiters by exposing the ML/NLP dataset to the recruiting team. Evaluation approaches must be understandable to various stakeholders, and useful for improving chatbot performance. The metric, nex-cv, uses negative examples in the evaluation of text classification, and fulfils three requirements. First, it is actionable: it can be used by non-developer staff. Second, it is not overly optimistic compared to human ratings, making it a fast method for comparing classifiers. Third, it allows model-agnostic comparison, making it useful for comparing systems despite implementation differences. We validate the metric based on seven recruitment-domain datasets in English and German over the course of one year.
Tasks	Chatbot, Text Classification
Published	2019-06-05
URL	https://arxiv.org/abs/1906.01910v1
PDF	https://arxiv.org/pdf/1906.01910v1.pdf
PWC	https://paperswithcode.com/paper/evaluation-and-improvement-of-chatbot-text
Repo	https://github.com/jobpal/nex-cv
Framework	none

Guiding High-Performance SAT Solvers with Unsat-Core Predictions


Title	Guiding High-Performance SAT Solvers with Unsat-Core Predictions
Authors	Daniel Selsam, Nikolaj Bjørner
Abstract	The NeuroSAT neural network architecture was recently introduced for predicting properties of propositional formulae. When trained to predict the satisfiability of toy problems, it was shown to find solutions and unsatisfiable cores on its own. However, the authors saw “no obvious path” to using the architecture to improve the state-of-the-art. In this work, we train a simplified NeuroSAT architecture to directly predict the unsatisfiable cores of real problems. We modify several high-performance SAT solvers to periodically replace their variable activity scores with NeuroSAT’s prediction of how likely the variables are to appear in an unsatisfiable core. The modified MiniSat solves 10% more problems on SAT-COMP 2018 within the standard 5,000 second timeout than the original does. The modified Glucose solves 11% more problems than the original, while the modified Z3 solves 6% more. The gains are even greater when the training is specialized for a specific distribution of problems; on a benchmark of hard problems from a scheduling domain, the modified Glucose solves 20% more problems than the original does within a one-hour timeout. Our results demonstrate that NeuroSAT can provide effective guidance to high-performance SAT solvers on real problems.
Tasks
Published	2019-03-12
URL	https://arxiv.org/abs/1903.04671v7
PDF	https://arxiv.org/pdf/1903.04671v7.pdf
PWC	https://paperswithcode.com/paper/neurocore-guiding-cdcl-with-unsat-core
Repo	https://github.com/dselsam/neurocore-public
Framework	tf

Stochastic first-order methods: non-asymptotic and computer-aided analyses via potential functions


Title	Stochastic first-order methods: non-asymptotic and computer-aided analyses via potential functions
Authors	Adrien Taylor, Francis Bach
Abstract	We provide a novel computer-assisted technique for systematically analyzing first-order methods for optimization. In contrast with previous works, the approach is particularly suited for handling sublinear convergence rates and stochastic oracles. The technique relies on semidefinite programming and potential functions. It allows simultaneously obtaining worst-case guarantees on the behavior of those algorithms, and assisting in choosing appropriate parameters for tuning their worst-case performances. The technique also benefits from comfortable tightness guarantees, meaning that unsatisfactory results can be improved only by changing the setting. We use the approach for analyzing deterministic and stochastic first-order methods under different assumptions on the nature of the stochastic noise. Among others, we treat unstructured noise with bounded variance, different noise models arising in over-parametrized expectation minimization problems, and randomized block-coordinate descent schemes.
Tasks
Published	2019-02-03
URL	https://arxiv.org/abs/1902.00947v4
PDF	https://arxiv.org/pdf/1902.00947v4.pdf
PWC	https://paperswithcode.com/paper/stochastic-first-order-methods-non-asymptotic
Repo	https://github.com/AdrienTaylor/Potential-functions-for-first-order-methods
Framework	none

BSP-Net: Generating Compact Meshes via Binary Space Partitioning


Title	BSP-Net: Generating Compact Meshes via Binary Space Partitioning
Authors	Zhiqin Chen, Andrea Tagliasacchi, Hao Zhang
Abstract	Polygonal meshes are ubiquitous in the digital 3D domain, yet they have only played a minor role in the deep learning revolution. Leading methods for learning generative models of shapes rely on implicit functions, and generate meshes only after expensive iso-surfacing routines. To overcome these challenges, we are inspired by a classical spatial data structure from computer graphics, Binary Space Partitioning (BSP), to facilitate 3D learning. The core ingredient of BSP is an operation for recursive subdivision of space to obtain convex sets. By exploiting this property, we devise BSP-Net, a network that learns to represent a 3D shape via convex decomposition. Importantly, BSP-Net is unsupervised since no convex shape decompositions are needed for training. The network is trained to reconstruct a shape using a set of convexes obtained from a BSP-tree built on a set of planes. The convexes inferred by BSP-Net can be easily extracted to form a polygon mesh, without any need for iso-surfacing. The generated meshes are compact (i.e., low-poly) and well suited to represent sharp geometry; they are guaranteed to be watertight and can be easily parameterized. We also show that the reconstruction quality by BSP-Net is competitive with state-of-the-art methods while using much fewer primitives. Code is available at https://github.com/czq142857/BSP-NET-original.
Tasks
Published	2019-11-16
URL	https://arxiv.org/abs/1911.06971v2
PDF	https://arxiv.org/pdf/1911.06971v2.pdf
PWC	https://paperswithcode.com/paper/bsp-net-generating-compact-meshes-via-binary
Repo	https://github.com/czq142857/BSP-NET-original
Framework	tf

Recognizing Musical Entities in User-generated Content


Title	Recognizing Musical Entities in User-generated Content
Authors	Lorenzo Porcaro, Horacio Saggion
Abstract	Recognizing Musical Entities is important for Music Information Retrieval (MIR) since it can improve the performance of several tasks such as music recommendation, genre classification or artist similarity. However, most entity recognition systems in the music domain have concentrated on formal texts (e.g. artists’ biographies, encyclopedic articles, etc.), ignoring rich and noisy user-generated content. In this work, we present a novel method to recognize musical entities in Twitter content generated by users following a classical music radio channel. Our approach takes advantage of both formal radio schedule and users’ tweets to improve entity recognition. We instantiate several machine learning algorithms to perform entity recognition combining task-specific and corpus-based features. We also show how to improve recognition results by jointly considering formal and user-generated content
Tasks	Information Retrieval, Music Information Retrieval
Published	2019-04-01
URL	http://arxiv.org/abs/1904.00648v1
PDF	http://arxiv.org/pdf/1904.00648v1.pdf
PWC	https://paperswithcode.com/paper/recognizing-musical-entities-in-user
Repo	https://github.com/LPorcaro/musicner
Framework	none

The Unfairness of Popularity Bias in Music Recommendation: A Reproducibility Study


Title	The Unfairness of Popularity Bias in Music Recommendation: A Reproducibility Study
Authors	Dominik Kowald, Markus Schedl, Elisabeth Lex
Abstract	Research has shown that recommender systems are typically biased towards popular items, which leads to less popular items being underrepresented in recommendations. The recent work of Abdollahpouri et al. in the context of movie recommendations has shown that this popularity bias leads to unfair treatment of both long-tail items as well as users with little interest in popular items. In this paper, we reproduce the analyses of Abdollahpouri et al. in the context of music recommendation. Specifically, we investigate three user groups from the LastFM music platform that are categorized based on how much their listening preferences deviate from the most popular music among all LastFM users in the dataset: (i) low-mainstream users, (ii) medium-mainstream users, and (iii) high-mainstream users. In line with Abdollahpouri et al., we find that state-of-the-art recommendation algorithms favor popular items also in the music domain. However, their proposed Group Average Popularity metric yields different results for LastFM than for the movie domain, presumably due to the larger number of available items (i.e., music artists) in the LastFM dataset we use. Finally, we compare the accuracy results of the recommendation algorithms for the three user groups and find that the low-mainstreaminess group significantly receives the worst recommendations.
Tasks	Recommendation Systems
Published	2019-12-10
URL	https://arxiv.org/abs/1912.04696v2
PDF	https://arxiv.org/pdf/1912.04696v2.pdf
PWC	https://paperswithcode.com/paper/the-unfairness-of-popularity-bias-in-music
Repo	https://github.com/domkowald/LFM1b-analyses
Framework	none

A Clustering Approach to Edge Controller Placement in Software Defined Networks with Cost Balancing


Title	A Clustering Approach to Edge Controller Placement in Software Defined Networks with Cost Balancing
Authors	Reza Soleymanifar, Amber Srivastava, Carolyn Beck, Srinivasa Salapaka
Abstract	In this work we introduce two novel deterministic annealing based clustering algorithms to address the problem of Edge Controller Placement (ECP) in wireless edge networks. These networks lie at the core of the fifth generation (5G) wireless systems and beyond. These algorithms, ECP-LL and ECP-LB, address the dominant leader-less and leader-based controller placement topologies and have linear computational complexity in terms of network size, maximum number of clusters and dimensionality of data. Each algorithm tries to place controllers close to edge node clusters and not far away from other controllers to maintain a reasonable balance between synchronization and delay costs. While the ECP problem can be conveniently expressed as a multi-objective mixed integer non-linear program (MINLP), our algorithms outperform state of art MINLP solver, BARON both in terms of accuracy and speed. Our proposed algorithms have the competitive edge of avoiding poor local minima through a Shannon entropy term in the clustering objective function. Most ECP algorithms are highly susceptible to poor local minima and greatly depend on initialization.
Tasks
Published	2019-12-05
URL	https://arxiv.org/abs/1912.02915v1
PDF	https://arxiv.org/pdf/1912.02915v1.pdf
PWC	https://paperswithcode.com/paper/a-clustering-approach-to-edge-controller
Repo	https://github.com/RezaSoleymanifar/DA-SDN
Framework	none

BitNet: Learning-Based Bit-Depth Expansion


Title	BitNet: Learning-Based Bit-Depth Expansion
Authors	Junyoung Byun, Kyujin Shim, Changick Kim
Abstract	Bit-depth is the number of bits for each color channel of a pixel in an image. Although many modern displays support unprecedented higher bit-depth to show more realistic and natural colors with a high dynamic range, most media sources are still in bit-depth of 8 or lower. Since insufficient bit-depth may generate annoying false contours or lose detailed visual appearance, bit-depth expansion (BDE) from low bit-depth (LBD) images to high bit-depth (HBD) images becomes more and more important. In this paper, we adopt a learning-based approach for BDE and propose a novel CNN-based bit-depth expansion network (BitNet) that can effectively remove false contours and restore visual details at the same time. We have carefully designed our BitNet based on an encoder-decoder architecture with dilated convolutions and a novel multi-scale feature integration. We have performed various experiments with four different datasets including MIT-Adobe FiveK, Kodak, ESPL v2, and TESTIMAGES, and our proposed BitNet has achieved state-of-the-art performance in terms of PSNR and SSIM among other existing BDE methods and famous CNN-based image processing networks. Unlike previous methods that separately process each color channel, we treat all RGB channels at once and have greatly improved color restoration. In addition, our network has shown the fastest computational speed in near real-time.
Tasks
Published	2019-10-10
URL	https://arxiv.org/abs/1910.04397v1
PDF	https://arxiv.org/pdf/1910.04397v1.pdf
PWC	https://paperswithcode.com/paper/bitnet-learning-based-bit-depth-expansion
Repo	https://github.com/kamkyu94/BitNet
Framework	tf

Multi-Pass Q-Networks for Deep Reinforcement Learning with Parameterised Action Spaces


Title	Multi-Pass Q-Networks for Deep Reinforcement Learning with Parameterised Action Spaces
Authors	Craig J. Bester, Steven D. James, George D. Konidaris
Abstract	Parameterised actions in reinforcement learning are composed of discrete actions with continuous action-parameters. This provides a framework for solving complex domains that require combining high-level actions with flexible control. The recent P-DQN algorithm extends deep Q-networks to learn over such action spaces. However, it treats all action-parameters as a single joint input to the Q-network, invalidating its theoretical foundations. We analyse the issues with this approach and propose a novel method, multi-pass deep Q-networks, or MP-DQN, to address them. We empirically demonstrate that MP-DQN significantly outperforms P-DQN and other previous algorithms in terms of data efficiency and converged policy performance on the Platform, Robot Soccer Goal, and Half Field Offense domains.
Tasks
Published	2019-05-10
URL	https://arxiv.org/abs/1905.04388v1
PDF	https://arxiv.org/pdf/1905.04388v1.pdf
PWC	https://paperswithcode.com/paper/multi-pass-q-networks-for-deep-reinforcement
Repo	https://github.com/cycraig/gym-platform
Framework	none

Training Artificial Neural Networks by Generalized Likelihood Ratio Method: Exploring Brain-like Learning to Improve Robustness


Title	Training Artificial Neural Networks by Generalized Likelihood Ratio Method: Exploring Brain-like Learning to Improve Robustness
Authors	Li Xiao, Yijie Peng, Jeff Hong, Zewu Ke, Shuhuai Yang
Abstract	In this work, we propose a generalized likelihood ratio method capable of training the artificial neural networks with some biological brain-like mechanisms,.e.g., (a) learning by the loss value, (b) learning via neurons with discontinuous activation and loss functions. The traditional back propagation method cannot train the artificial neural networks with aforementioned brain-like learning mechanisms. Numerical results show that the robustness of various artificial neural networks trained by the new method is significantly improved when the input data is affected by both the natural noises and adversarial attacks. Code is available: \url{https://github.com/LX-doctorAI/GLR_ADV} .
Tasks
Published	2019-01-31
URL	https://arxiv.org/abs/1902.00358v2
PDF	https://arxiv.org/pdf/1902.00358v2.pdf
PWC	https://paperswithcode.com/paper/training-artificial-neural-networks-by
Repo	https://github.com/LX-doctorAI/GLR_ADV
Framework	pytorch

DeepRec: An Open-source Toolkit for Deep Learning based Recommendation


Title	DeepRec: An Open-source Toolkit for Deep Learning based Recommendation
Authors	Shuai Zhang, Yi Tay, Lina Yao, Bin Wu, Aixin Sun
Abstract	Deep learning based recommender systems have been extensively explored in recent years. However, the large number of models proposed each year poses a big challenge for both researchers and practitioners in reproducing the results for further comparisons. Although a portion of papers provides source code, they adopted different programming languages or different deep learning packages, which also raises the bar in grasping the ideas. To alleviate this problem, we released the open source project: \textbf{DeepRec}. In this toolkit, we have implemented a number of deep learning based recommendation algorithms using Python and the widely used deep learning package - Tensorflow. Three major recommendation scenarios: rating prediction, top-N recommendation (item ranking) and sequential recommendation, were considered. Meanwhile, DeepRec maintains good modularity and extensibility to easily incorporate new models into the framework. It is distributed under the terms of the GNU General Public License. The source code is available at github: \url{https://github.com/cheungdaven/DeepRec}
Tasks	Recommendation Systems
Published	2019-05-25
URL	https://arxiv.org/abs/1905.10536v1
PDF	https://arxiv.org/pdf/1905.10536v1.pdf
PWC	https://paperswithcode.com/paper/deeprec-an-open-source-toolkit-for-deep
Repo	https://github.com/cheungdaven/DeepRec
Framework	tf