April 1, 2020

3231 words 16 mins read

Paper Group ANR 421

Inferring Semantic Information with 3D Neural Scene Representations. Explosive Proofs of Mathematical Truths. Hypernetwork approach to generating point clouds. TableQnA: Answering List Intent Queries With Web Tables. A Nepali Rule Based Stemmer and its performance on different NLP applications. Bayesian System ID: Optimal management of parameter, m …

Inferring Semantic Information with 3D Neural Scene Representations


Title	Inferring Semantic Information with 3D Neural Scene Representations
Authors	Amit Kohli, Vincent Sitzmann, Gordon Wetzstein
Abstract	Biological vision infers multi-modal 3D representations that support reasoning about scene properties such as materials, appearance, affordance, and semantics in 3D. These rich representations enable us humans, for example, to acquire new skills, such as the learning of a new semantic class, with extremely limited supervision. Motivated by this ability of biological vision, we demonstrate that 3D-structure-aware representation learning leads to multi-modal representations that enable 3D semantic segmentation with extremely limited, 2D-only supervision. Building on emerging neural scene representations, which have been developed for modeling the shape and appearance of 3D scenes supervised exclusively by posed 2D images, we are first to demonstrate a representation that jointly encodes shape, appearance, and semantics in a 3D-structure-aware manner. Surprisingly, we find that only a few tens of labeled 2D segmentation masks are required to achieve dense 3D semantic segmentation using a semi-supervised learning strategy. We explore two novel applications for our semantically aware neural scene representation: 3D novel view and semantic label synthesis given only a single input RGB image or 2D label mask, as well as 3D interpolation of appearance and semantics.
Tasks	3D Semantic Segmentation, Representation Learning, Semantic Segmentation
Published	2020-03-28
URL	https://arxiv.org/abs/2003.12673v1
PDF	https://arxiv.org/pdf/2003.12673v1.pdf
PWC	https://paperswithcode.com/paper/inferring-semantic-information-with-3d-neural
Repo
Framework

Explosive Proofs of Mathematical Truths


Title	Explosive Proofs of Mathematical Truths
Authors	Scott Viteri, Simon DeDeo
Abstract	Mathematical proofs are both paradigms of certainty and some of the most explicitly-justified arguments that we have in the cultural record. Their very explicitness, however, leads to a paradox, because their probability of error grows exponentially as the argument expands. Here we show that under a cognitively-plausible belief formation mechanism that combines deductive and abductive reasoning, mathematical arguments can undergo what we call an epistemic phase transition: a dramatic and rapidly-propagating jump from uncertainty to near-complete confidence at reasonable levels of claim-to-claim error rates. To show this, we analyze an unusual dataset of forty-eight machine-aided proofs from the formalized reasoning system Coq, including major theorems ranging from ancient to 21st Century mathematics, along with four hand-constructed cases from Euclid, Apollonius, Spinoza, and Andrew Wiles. Our results bear both on recent work in the history and philosophy of mathematics, and on a question, basic to cognitive science, of how we form beliefs, and justify them to others.
Tasks	Mathematical Proofs
Published	2020-03-31
URL	https://arxiv.org/abs/2004.00055v1
PDF	https://arxiv.org/pdf/2004.00055v1.pdf
PWC	https://paperswithcode.com/paper/explosive-proofs-of-mathematical-truths
Repo
Framework

Hypernetwork approach to generating point clouds


Title	Hypernetwork approach to generating point clouds
Authors	Przemysław Spurek, Sebastian Winczowski, Jacek Tabor, Maciej Zamorski, Maciej Zięba, Tomasz Trzciński
Abstract	In this work, we propose a novel method for generating 3D point clouds that leverage properties of hyper networks. Contrary to the existing methods that learn only the representation of a 3D object, our approach simultaneously finds a representation of the object and its 3D surface. The main idea of our HyperCloud method is to build a hyper network that returns weights of a particular neural network (target network) trained to map points from a uniform unit ball distribution into a 3D shape. As a consequence, a particular 3D shape can be generated using point-by-point sampling from the assumed prior distribution and transforming sampled points with the target network. Since the hyper network is based on an auto-encoder architecture trained to reconstruct realistic 3D shapes, the target network weights can be considered a parametrization of the surface of a 3D shape, and not a standard representation of point cloud usually returned by competitive approaches. The proposed architecture allows finding mesh-based representation of 3D objects in a generative manner while providing point clouds en pair in quality with the state-of-the-art methods.
Tasks	Generating 3D Point Clouds
Published	2020-02-10
URL	https://arxiv.org/abs/2003.00802v1
PDF	https://arxiv.org/pdf/2003.00802v1.pdf
PWC	https://paperswithcode.com/paper/hypernetwork-approach-to-generating-point
Repo
Framework

TableQnA: Answering List Intent Queries With Web Tables


Title	TableQnA: Answering List Intent Queries With Web Tables
Authors	Kaushik Chakrabarti, Zhimin Chen, Siamak Shakeri, Guihong Cao, Surajit Chaudhuri
Abstract	The web contains a vast corpus of HTML tables. They can be used to provide direct answers to many web queries. We focus on answering two classes of queries with those tables: those seeking lists of entities (e.g., `cities in california') and those seeking superlative entities (e.g.,` largest city in california’). The main challenge is to achieve high precision with significant coverage. Existing approaches train machine learning models to select the answer from the candidates; they rely on textual match features between the query and the content of the table along with features capturing table quality/importance. These features alone are inadequate for achieving the above goals. Our main insight is that we can improve precision by (i) first extracting intent (structured information) from the query for the above query classes and (ii) then performing structure-aware matching (instead of just textual matching) between the extracted intent and the candidates to select the answer. We model (i) as a sequence tagging task. We leverage state-of-the-art deep neural network models with word embeddings. The model requires large scale training data which is expensive to obtain via manual labeling; we therefore develop a novel method to automatically generate the training data. For (ii), we develop novel features to compute structure-aware match and train a machine learning model. Our experiments on real-life web search queries show that (i) our intent extractor for list and superlative intent queries has significantly higher precision and coverage compared with baseline approaches and (ii) our table answer selector significantly outperforms the state-of-the-art baseline approach. This technology has been used in production by Microsoft’s Bing search engine since 2016.
Tasks	Word Embeddings
Published	2020-01-10
URL	https://arxiv.org/abs/2001.04828v1
PDF	https://arxiv.org/pdf/2001.04828v1.pdf
PWC	https://paperswithcode.com/paper/tableqna-answering-list-intent-queries-with
Repo
Framework

A Nepali Rule Based Stemmer and its performance on different NLP applications


Title	A Nepali Rule Based Stemmer and its performance on different NLP applications
Authors	Pravesh Koirala, Aman Shakya
Abstract	Stemming is an integral part of Natural Language Processing (NLP). It’s a preprocessing step in almost every NLP application. Arguably, the most important usage of stemming is in Information Retrieval (IR). While there are lots of work done on stemming in languages like English, Nepali stemming has only a few works. This study focuses on creating a Rule Based stemmer for Nepali text. Specifically, it is an affix stripping system that identifies two different class of suffixes in Nepali grammar and strips them separately. Only a single negativity prefix (Na) is identified and stripped. This study focuses on a number of techniques like exception word identification, morphological normalization and word transformation to increase stemming performance. The stemmer is tested intrinsically using Paice’s method and extrinsically on a basic tf-idf based IR system and an elementary news topic classifier using Multinomial Naive Bayes Classifier. The difference in performance of these systems with and without using the stemmer is analysed.
Tasks	Information Retrieval
Published	2020-02-23
URL	https://arxiv.org/abs/2002.09901v1
PDF	https://arxiv.org/pdf/2002.09901v1.pdf
PWC	https://paperswithcode.com/paper/a-nepali-rule-based-stemmer-and-its
Repo
Framework

Bayesian System ID: Optimal management of parameter, model, and measurement uncertainty


Title	Bayesian System ID: Optimal management of parameter, model, and measurement uncertainty
Authors	Nicholas Galioto, Alex Gorodetsky
Abstract	We evaluate the robustness of a probabilistic formulation of system identification (ID) to sparse, noisy, and indirect data. Specifically, we compare estimators of future system behavior derived from the Bayesian posterior of a learning problem to several commonly used least squares-based optimization objectives used in system ID. Our comparisons indicate that the log posterior has improved geometric properties compared with the objective function surfaces of traditional methods that include differentially constrained least squares and least squares reconstructions of discrete time steppers like dynamic mode decomposition (DMD). These properties allow it to be both more sensitive to new data and less affected by multiple minima — overall yielding a more robust approach. Our theoretical results indicate that least squares and regularized least squares methods like dynamic mode decomposition and sparse identification of nonlinear dynamics (SINDy) can be derived from the probabilistic formulation by assuming noiseless measurements. We also analyze the computational complexity of a Gaussian filter-based approximate marginal Markov Chain Monte Carlo scheme that we use to obtain the Bayesian posterior for both linear and nonlinear problems. We then empirically demonstrate that obtaining the marginal posterior of the parameter dynamics and making predictions by extracting optimal estimators (e.g., mean, median, mode) yields orders of magnitude improvement over the aforementioned approaches. We attribute this performance to the fact that the Bayesian approach captures parameter, model, and measurement uncertainties, whereas the other methods typically neglect at least one type of uncertainty.
Tasks
Published	2020-03-04
URL	https://arxiv.org/abs/2003.02359v1
PDF	https://arxiv.org/pdf/2003.02359v1.pdf
PWC	https://paperswithcode.com/paper/bayesian-system-id-optimal-management-of
Repo
Framework

Do ML Experts Discuss Explainability for AI Systems? A discussion case in the industry for a domain-specific solution


Title	Do ML Experts Discuss Explainability for AI Systems? A discussion case in the industry for a domain-specific solution
Authors	Juliana Jansen Ferreira, Mateus de Souza Monteiro
Abstract	The application of Artificial Intelligence (AI) tools in different domains are becoming mandatory for all companies wishing to excel in their industries. One major challenge for a successful application of AI is to combine the machine learning (ML) expertise with the domain knowledge to have the best results applying AI tools. Domain specialists have an understanding of the data and how it can impact their decisions. ML experts have the ability to use AI-based tools dealing with large amounts of data and generating insights for domain experts. But without a deep understanding of the data, ML experts are not able to tune their models to get optimal results for a specific domain. Therefore, domain experts are key users for ML tools and the explainability of those AI tools become an essential feature in that context. There are a lot of efforts to research AI explainability for different contexts, users and goals. In this position paper, we discuss interesting findings about how ML experts can express concerns about AI explainability while defining features of an ML tool to be developed for a specific domain. We analyze data from two brainstorm sessions done to discuss the functionalities of an ML tool to support geoscientists (domain experts) on analyzing seismic data (domain-specific data) with ML resources.
Tasks
Published	2020-02-27
URL	https://arxiv.org/abs/2002.12450v1
PDF	https://arxiv.org/pdf/2002.12450v1.pdf
PWC	https://paperswithcode.com/paper/do-ml-experts-discuss-explainability-for-ai
Repo
Framework

Towards High Performance Java-based Deep Learning Frameworks


Title	Towards High Performance Java-based Deep Learning Frameworks
Authors	Athanasios Stratikopoulos, Juan Fumero, Zoran Sevarac, Christos Kotselidis
Abstract	The advent of modern cloud services along with the huge volume of data produced on a daily basis, have set the demand for fast and efficient data processing. This demand is common among numerous application domains, such as deep learning, data mining, and computer vision. Prior research has focused on employing hardware accelerators as a means to overcome this inefficiency. This trend has driven software development to target heterogeneous execution, and several modern computing systems have incorporated a mixture of diverse computing components, including GPUs and FPGAs. However, the specialization of the applications’ code for heterogeneous execution is not a trivial task, as it requires developers to have hardware expertise in order to obtain high performance. The vast majority of the existing deep learning frameworks that support heterogeneous acceleration, rely on the implementation of wrapper calls from a high-level programming language to a low-level accelerator backend, such as OpenCL, CUDA or HLS. In this paper we have employed TornadoVM, a state-of-the-art heterogeneous programming framework to transparently accelerate Deep Netts; a Java-based deep learning framework. Our initial results demonstrate up to 8x performance speedup when executing the back propagation process of the network’s training on AMD GPUs against the sequential execution of the original Deep Netts framework.
Tasks
Published	2020-01-13
URL	https://arxiv.org/abs/2001.04206v1
PDF	https://arxiv.org/pdf/2001.04206v1.pdf
PWC	https://paperswithcode.com/paper/towards-high-performance-java-based-deep
Repo
Framework

Speech Emotion Recognition using Support Vector Machine


Title	Speech Emotion Recognition using Support Vector Machine
Authors	Manas Jain, Shruthi Narayan, Pratibha Balaji, Bharath K P, Abhijit Bhowmick, Karthik R, Rajesh Kumar Muthu
Abstract	In this project, we aim to classify the speech taken as one of the four emotions namely, sadness, anger, fear and happiness. The samples that have been taken to complete this project are taken from Linguistic Data Consortium (LDC) and UGA database. The important characteristics determined from the samples are energy, pitch, MFCC coefficients, LPCC coefficients and speaker rate. The classifier used to classify these emotional states is Support Vector Machine (SVM) and this is done using two classification strategies: One against All (OAA) and Gender Dependent Classification. Furthermore, a comparative analysis has been conducted between the two and LPCC and MFCC algorithms as well.
Tasks	Emotion Recognition, Speech Emotion Recognition
Published	2020-02-03
URL	https://arxiv.org/abs/2002.07590v1
PDF	https://arxiv.org/pdf/2002.07590v1.pdf
PWC	https://paperswithcode.com/paper/speech-emotion-recognition-using-support
Repo
Framework

Segmentation and Recovery of Superquadric Models using Convolutional Neural Networks


Title	Segmentation and Recovery of Superquadric Models using Convolutional Neural Networks
Authors	Jaka Šircelj, Tim Oblak, Klemen Grm, Uroš Petković, Aleš Jaklič, Peter Peer, Vitomir Štruc, Franc Solina
Abstract	In this paper we address the problem of representing 3D visual data with parameterized volumetric shape primitives. Specifically, we present a (two-stage) approach built around convolutional neural networks (CNNs) capable of segmenting complex depth scenes into the simpler geometric structures that can be represented with superquadric models. In the first stage, our approach uses a Mask RCNN model to identify superquadric-like structures in depth scenes and then fits superquadric models to the segmented structures using a specially designed CNN regressor. Using our approach we are able to describe complex structures with a small number of interpretable parameters. We evaluated the proposed approach on synthetic as well as real-world depth data and show that our solution does not only result in competitive performance in comparison to the state-of-the-art, but is able to decompose scenes into a number of superquadric models at a fraction of the time required by competing approaches. We make all data and models used in the paper available from https://lmi.fe.uni-lj.si/en/research/resources/sq-seg.
Tasks
Published	2020-01-28
URL	https://arxiv.org/abs/2001.10504v1
PDF	https://arxiv.org/pdf/2001.10504v1.pdf
PWC	https://paperswithcode.com/paper/segmentation-and-recovery-of-superquadric
Repo
Framework

Predicting Unplanned Readmissions with Highly Unstructured Data


Title	Predicting Unplanned Readmissions with Highly Unstructured Data
Authors	Constanza Fierro, Jorge Pérez, Javier Mora
Abstract	Deep learning techniques have been successfully applied to predict unplanned readmissions of patients in medical centers. The training data for these models is usually based on historical medical records that contain a significant amount of free-text from admission reports, referrals, exam notes, etc. Most of the models proposed so far are tailored to English text data and assume that electronic medical records follow standards common in developed countries. These two characteristics make them difficult to apply in developing countries that do not necessarily follow international standards for registering patient information, or that store text information in languages other than English. In this paper we propose a deep learning architecture for predicting unplanned readmissions that consumes data that is significantly less structured compared with previous models in the literature. We use it to present the first results for this task in a large clinical dataset that mainly contains Spanish text data. The dataset is composed of almost 10 years of records in a Chilean medical center. On this dataset, our model achieves results that are comparable to some of the most recent results obtained in US medical centers for the same task (0.76 AUROC).
Tasks
Published	2020-03-19
URL	https://arxiv.org/abs/2003.11622v1
PDF	https://arxiv.org/pdf/2003.11622v1.pdf
PWC	https://paperswithcode.com/paper/predicting-unplanned-readmissions-with-highly
Repo
Framework

Black-Box Saliency Map Generation Using Bayesian Optimisation


Title	Black-Box Saliency Map Generation Using Bayesian Optimisation
Authors	Mamuku Mokuwe, Michael Burke, Anna Sergeevna Bosman
Abstract	Saliency maps are often used in computer vision to provide intuitive interpretations of what input regions a model has used to produce a specific prediction. A number of approaches to saliency map generation are available, but most require access to model parameters. This work proposes an approach for saliency map generation for black-box models, where no access to model parameters is available, using a Bayesian optimisation sampling method. The approach aims to find the global salient image region responsible for a particular (black-box) model’s prediction. This is achieved by a sampling-based approach to model perturbations that seeks to localise salient regions of an image to the black-box model. Results show that the proposed approach to saliency map generation outperforms grid-based perturbation approaches, and performs similarly to gradient-based approaches which require access to model parameters.
Tasks	Bayesian Optimisation
Published	2020-01-30
URL	https://arxiv.org/abs/2001.11366v1
PDF	https://arxiv.org/pdf/2001.11366v1.pdf
PWC	https://paperswithcode.com/paper/black-box-saliency-map-generation-using
Repo
Framework

Minimax Optimal Estimation of KL Divergence for Continuous Distributions


Title	Minimax Optimal Estimation of KL Divergence for Continuous Distributions
Authors	Puning Zhao, Lifeng Lai
Abstract	Estimating Kullback-Leibler divergence from identical and independently distributed samples is an important problem in various domains. One simple and effective estimator is based on the k nearest neighbor distances between these samples. In this paper, we analyze the convergence rates of the bias and variance of this estimator. Furthermore, we derive a lower bound of the minimax mean square error and show that kNN method is asymptotically rate optimal.
Tasks
Published	2020-02-26
URL	https://arxiv.org/abs/2002.11599v1
PDF	https://arxiv.org/pdf/2002.11599v1.pdf
PWC	https://paperswithcode.com/paper/minimax-optimal-estimation-of-kl-divergence
Repo
Framework

Rapid Adaptation of BERT for Information Extraction on Domain-Specific Business Documents


Title	Rapid Adaptation of BERT for Information Extraction on Domain-Specific Business Documents
Authors	Ruixue Zhang, Wei Yang, Luyun Lin, Zhengkai Tu, Yuqing Xie, Zihang Fu, Yuhao Xie, Luchen Tan, Kun Xiong, Jimmy Lin
Abstract	Techniques for automatically extracting important content elements from business documents such as contracts, statements, and filings have the potential to make business operations more efficient. This problem can be formulated as a sequence labeling task, and we demonstrate the adaption of BERT to two types of business documents: regulatory filings and property lease agreements. There are aspects of this problem that make it easier than “standard” information extraction tasks and other aspects that make it more difficult, but on balance we find that modest amounts of annotated data (less than 100 documents) are sufficient to achieve reasonable accuracy. We integrate our models into an end-to-end cloud platform that provides both an easy-to-use annotation interface as well as an inference interface that allows users to upload documents and inspect model outputs.
Tasks
Published	2020-02-05
URL	https://arxiv.org/abs/2002.01861v1
PDF	https://arxiv.org/pdf/2002.01861v1.pdf
PWC	https://paperswithcode.com/paper/rapid-adaptation-of-bert-for-information
Repo
Framework

Summarizing the performances of a background subtraction algorithm measured on several videos


Title	Summarizing the performances of a background subtraction algorithm measured on several videos
Authors	Sébastien Piérard, Marc Van Droogenbroeck
Abstract	There exist many background subtraction algorithms to detect motion in videos. To help comparing them, datasets with ground-truth data such as CDNET or LASIESTA have been proposed. These datasets organize videos in categories that represent typical challenges for background subtraction. The evaluation procedure promoted by their authors consists in measuring performance indicators for each video separately and to average them hierarchically, within a category first, then between categories, a procedure which we name “summarization”. While the summarization by averaging performance indicators is a valuable effort to standardize the evaluation procedure, it has no theoretical justification and it breaks the intrinsic relationships between summarized indicators. This leads to interpretation inconsistencies. In this paper, we present a theoretical approach to summarize the performances for multiple videos that preserves the relationships between performance indicators. In addition, we give formulas and an algorithm to calculate summarized performances. Finally, we showcase our observations on CDNET 2014.
Tasks
Published	2020-02-13
URL	https://arxiv.org/abs/2002.05654v1
PDF	https://arxiv.org/pdf/2002.05654v1.pdf
PWC	https://paperswithcode.com/paper/summarizing-the-performances-of-a-background
Repo
Framework