October 16, 2019

2839 words 14 mins read

Paper Group ANR 1000

3D-LaneNet: End-to-End 3D Multiple Lane Detection. WEST: Word Encoded Sequence Transducers. Replay spoofing detection system for automatic speaker verification using multi-task learning of noise classes. Rejoinder for “Probabilistic Integration: A Role in Statistical Computation?". SEGA: Variance Reduction via Gradient Sketching. Labeling Panoramas …

3D-LaneNet: End-to-End 3D Multiple Lane Detection


Title	3D-LaneNet: End-to-End 3D Multiple Lane Detection
Authors	Noa Garnett, Rafi Cohen, Tomer Pe’er, Roee Lahav, Dan Levi
Abstract	We introduce a network that directly predicts the 3D layout of lanes in a road scene from a single image. This work marks a first attempt to address this task with on-board sensing without assuming a known constant lane width or relying on pre-mapped environments. Our network architecture, 3D-LaneNet, applies two new concepts: intra-network inverse-perspective mapping (IPM) and anchor-based lane representation. The intra-network IPM projection facilitates a dual-representation information flow in both regular image-view and top-view. An anchor-per-column output representation enables our end-to-end approach which replaces common heuristics such as clustering and outlier rejection, casting lane estimation as an object detection problem. In addition, our approach explicitly handles complex situations such as lane merges and splits. Results are shown on two new 3D lane datasets, a synthetic and a real one. For comparison with existing methods, we test our approach on the image-only tuSimple lane detection benchmark, achieving performance competitive with state-of-the-art.
Tasks	Lane Detection, Object Detection
Published	2018-11-26
URL	https://arxiv.org/abs/1811.10203v3
PDF	https://arxiv.org/pdf/1811.10203v3.pdf
PWC	https://paperswithcode.com/paper/3d-lanenet-end-to-end-3d-multiple-lane
Repo
Framework

WEST: Word Encoded Sequence Transducers


Title	WEST: Word Encoded Sequence Transducers
Authors	Ehsan Variani, Ananda Theertha Suresh, Mitchel Weintraub
Abstract	Most of the parameters in large vocabulary models are used in embedding layer to map categorical features to vectors and in softmax layer for classification weights. This is a bottle-neck in memory constraint on-device training applications like federated learning and on-device inference applications like automatic speech recognition (ASR). One way of compressing the embedding and softmax layers is to substitute larger units such as words with smaller sub-units such as characters. However, often the sub-unit models perform poorly compared to the larger unit models. We propose WEST, an algorithm for encoding categorical features and output classes with a sequence of random or domain dependent sub-units and demonstrate that this transduction can lead to significant compression without compromising performance. WEST bridges the gap between larger unit and sub-unit models and can be interpreted as a MaxEnt model over sub-unit features, which can be of independent interest.
Tasks	Speech Recognition
Published	2018-11-20
URL	http://arxiv.org/abs/1811.08417v1
PDF	http://arxiv.org/pdf/1811.08417v1.pdf
PWC	https://paperswithcode.com/paper/west-word-encoded-sequence-transducers
Repo
Framework

Replay spoofing detection system for automatic speaker verification using multi-task learning of noise classes


Title	Replay spoofing detection system for automatic speaker verification using multi-task learning of noise classes
Authors	Hye-Jin Shim, Jee-weon Jung, Hee-Soo Heo, Sunghyun Yoon, Ha-Jin Yu
Abstract	In this paper, we propose a replay attack spoofing detection system for automatic speaker verification using multitask learning of noise classes. We define the noise that is caused by the replay attack as replay noise. We explore the effectiveness of training a deep neural network simultaneously for replay attack spoofing detection and replay noise classification. The multi-task learning includes classifying the noise of playback devices, recording environments, and recording devices as well as the spoofing detection. Each of the three types of the noise classes also includes a genuine class. The experiment results on the ASVspoof2017 datasets demonstrate that the performance of our proposed system is improved by 30% relatively on the evaluation set.
Tasks	Multi-Task Learning, Speaker Verification
Published	2018-08-29
URL	http://arxiv.org/abs/1808.09638v4
PDF	http://arxiv.org/pdf/1808.09638v4.pdf
PWC	https://paperswithcode.com/paper/replay-spoofing-detection-system-for
Repo
Framework

Rejoinder for “Probabilistic Integration: A Role in Statistical Computation?”


Title	Rejoinder for “Probabilistic Integration: A Role in Statistical Computation?”
Authors	Francois-Xavier Briol, Chris J. Oates, Mark Girolami, Michael A. Osborne, Dino Sejdinovic
Abstract	This article is the rejoinder for the paper “Probabilistic Integration: A Role in Statistical Computation?” to appear in Statistical Science with discussion. We would first like to thank the reviewers and many of our colleagues who helped shape this paper, the editor for selecting our paper for discussion, and of course all of the discussants for their thoughtful, insightful and constructive comments. In this rejoinder, we respond to some of the points raised by the discussants and comment further on the fundamental questions underlying the paper: (i) Should Bayesian ideas be used in numerical analysis?, and (ii) If so, what role should such approaches have in statistical computation?
Tasks
Published	2018-11-26
URL	http://arxiv.org/abs/1811.10275v1
PDF	http://arxiv.org/pdf/1811.10275v1.pdf
PWC	https://paperswithcode.com/paper/rejoinder-for-probabilistic-integration-a
Repo
Framework

SEGA: Variance Reduction via Gradient Sketching


Title	SEGA: Variance Reduction via Gradient Sketching
Authors	Filip Hanzely, Konstantin Mishchenko, Peter Richtarik
Abstract	We propose a randomized first order optimization method–SEGA (SkEtched GrAdient method)– which progressively throughout its iterations builds a variance-reduced estimate of the gradient from random linear measurements (sketches) of the gradient obtained from an oracle. In each iteration, SEGA updates the current estimate of the gradient through a sketch-and-project operation using the information provided by the latest sketch, and this is subsequently used to compute an unbiased estimate of the true gradient through a random relaxation procedure. This unbiased estimate is then used to perform a gradient step. Unlike standard subspace descent methods, such as coordinate descent, SEGA can be used for optimization problems with a non-separable proximal term. We provide a general convergence analysis and prove linear convergence for strongly convex objectives. In the special case of coordinate sketches, SEGA can be enhanced with various techniques such as importance sampling, minibatching and acceleration, and its rate is up to a small constant factor identical to the best-known rate of coordinate descent.
Tasks
Published	2018-09-09
URL	http://arxiv.org/abs/1809.03054v2
PDF	http://arxiv.org/pdf/1809.03054v2.pdf
PWC	https://paperswithcode.com/paper/sega-variance-reduction-via-gradient
Repo
Framework

Labeling Panoramas with Spherical Hourglass Networks


Title	Labeling Panoramas with Spherical Hourglass Networks
Authors	Carlos Esteves, Kostas Daniilidis, Ameesh Makadia
Abstract	With the recent proliferation of consumer-grade 360{\deg} cameras, it is worth revisiting visual perception challenges with spherical cameras given the potential benefit of their global field of view. To this end we introduce a spherical convolutional hourglass network (SCHN) for the dense labeling on the sphere. The SCHN is invariant to camera orientation (lifting the usual requirement for `upright’ panoramic images), and its design is scalable for larger practical datasets. Initial experiments show promising results on a spherical semantic segmentation task. \|
Tasks	Semantic Segmentation
Published	2018-09-06
URL	http://arxiv.org/abs/1809.02123v1
PDF	http://arxiv.org/pdf/1809.02123v1.pdf
PWC	https://paperswithcode.com/paper/labeling-panoramas-with-spherical-hourglass
Repo
Framework

Hierarchical Attention: What Really Counts in Various NLP Tasks


Title	Hierarchical Attention: What Really Counts in Various NLP Tasks
Authors	Zehao Dou, Zhihua Zhang
Abstract	Attention mechanisms in sequence to sequence models have shown great ability and wonderful performance in various natural language processing (NLP) tasks, such as sentence embedding, text generation, machine translation, machine reading comprehension, etc. Unfortunately, existing attention mechanisms only learn either high-level or low-level features. In this paper, we think that the lack of hierarchical mechanisms is a bottleneck in improving the performance of the attention mechanisms, and propose a novel Hierarchical Attention Mechanism (Ham) based on the weighted sum of different layers of a multi-level attention. Ham achieves a state-of-the-art BLEU score of 0.26 on Chinese poem generation task and a nearly 6.5% averaged improvement compared with the existing machine reading comprehension models such as BIDAF and Match-LSTM. Furthermore, our experiments and theorems reveal that Ham has greater generalization and representation ability than existing attention mechanisms.
Tasks	Machine Reading Comprehension, Machine Translation, Reading Comprehension, Sentence Embedding, Text Generation
Published	2018-08-10
URL	http://arxiv.org/abs/1808.03728v1
PDF	http://arxiv.org/pdf/1808.03728v1.pdf
PWC	https://paperswithcode.com/paper/hierarchical-attention-what-really-counts-in
Repo
Framework

Toward Fast and Accurate Neural Discourse Segmentation


Title	Toward Fast and Accurate Neural Discourse Segmentation
Authors	Yizhong Wang, Sujian Li, Jingfeng Yang
Abstract	Discourse segmentation, which segments texts into Elementary Discourse Units, is a fundamental step in discourse analysis. Previous discourse segmenters rely on complicated hand-crafted features and are not practical in actual use. In this paper, we propose an end-to-end neural segmenter based on BiLSTM-CRF framework. To improve its accuracy, we address the problem of data insufficiency by transferring a word representation model that is trained on a large corpus. We also propose a restricted self-attention mechanism in order to capture useful information within a neighborhood. Experiments on the RST-DT corpus show that our model is significantly faster than previous methods, while achieving new state-of-the-art performance.
Tasks
Published	2018-08-28
URL	http://arxiv.org/abs/1808.09147v1
PDF	http://arxiv.org/pdf/1808.09147v1.pdf
PWC	https://paperswithcode.com/paper/toward-fast-and-accurate-neural-discourse
Repo
Framework

Online Learning for Non-Stationary A/B Tests


Title	Online Learning for Non-Stationary A/B Tests
Authors	Andrés Muñoz Medina, Sergei Vassilvitskii, Dong Yin
Abstract	The rollout of new versions of a feature in modern applications is a manual multi-stage process, as the feature is released to ever larger groups of users, while its performance is carefully monitored. This kind of A/B testing is ubiquitous, but suboptimal, as the monitoring requires heavy human intervention, is not guaranteed to capture consistent, but short-term fluctuations in performance, and is inefficient, as better versions take a long time to reach the full population. In this work we formulate this question as that of expert learning, and give a new algorithm Follow-The-Best-Interval, FTBI, that works in dynamic, non-stationary environments. Our approach is practical, simple, and efficient, and has rigorous guarantees on its performance. Finally, we perform a thorough evaluation on synthetic and real world datasets and show that our approach outperforms current state-of-the-art methods.
Tasks
Published	2018-02-14
URL	http://arxiv.org/abs/1802.05315v2
PDF	http://arxiv.org/pdf/1802.05315v2.pdf
PWC	https://paperswithcode.com/paper/online-learning-for-non-stationary-ab-tests
Repo
Framework

Segmentation of both Diseased and Healthy Skin from Clinical Photographs in a Primary Care Setting


Title	Segmentation of both Diseased and Healthy Skin from Clinical Photographs in a Primary Care Setting
Authors	Noel C. F. Codella, Daren Anderson, Tyler Philips, Anthony Porto, Kevin Massey, Jane Snowdon, Rogerio Feris, John Smith
Abstract	This work presents the first segmentation study of both diseased and healthy skin in standard camera photographs from a clinical environment. Challenges arise from varied lighting conditions, skin types, backgrounds, and pathological states. For study, 400 clinical photographs (with skin segmentation masks) representing various pathological states of skin are retrospectively collected from a primary care network. 100 images are used for training and fine-tuning, and 300 are used for evaluation. This distribution between training and test partitions is chosen to reflect the difficulty in amassing large quantities of labeled data in this domain. A deep learning approach is used, and 3 public segmentation datasets of healthy skin are collected to study the potential benefits of pre-training. Two variants of U-Net are evaluated: U-Net and Dense Residual U-Net. We find that Dense Residual U-Nets have a 7.8% improvement in Jaccard, compared to classical U-Net architectures (0.55 vs. 0.51 Jaccard), for direct transfer, where fine-tuning data is not utilized. However, U-Net outperforms Dense Residual U-Net for both direct training (0.83 vs. 0.80) and fine-tuning (0.89 vs. 0.88). The stark performance improvement with fine-tuning compared to direct transfer and direct training emphasizes both the need for adequate representative data of diseased skin, and the utility of other publicly available data sources for this task.
Tasks
Published	2018-04-16
URL	http://arxiv.org/abs/1804.05944v2
PDF	http://arxiv.org/pdf/1804.05944v2.pdf
PWC	https://paperswithcode.com/paper/segmentation-of-both-diseased-and-healthy
Repo
Framework

Risk-Stratify: Confident Stratification Of Patients Based On Risk


Title	Risk-Stratify: Confident Stratification Of Patients Based On Risk
Authors	Kartik Ahuja, Mihaela van der Schaar
Abstract	A clinician desires to use a risk-stratification method that achieves confident risk-stratification - the risk estimates of the different patients reflect the true risks with a high probability. This allows him/her to use these risks to make accurate predictions about prognosis and decisions about screening, treatments for the current patient. We develop Risk-stratify - a two phase algorithm that is designed to achieve confident risk-stratification. In the first phase, we grow a tree to partition the covariate space. Each node in the tree is split using statistical tests that determine if the risks of the child nodes are different or not. The choice of the statistical tests depends on whether the data is censored (Log-rank test) or not (U-test). The set of the leaves of the tree form a partition. The risk distribution of patients that belong to a leaf is different from the sibling leaf but not the rest of the leaves. Therefore, some of the leaves that have similar underlying risks are incorrectly specified to have different risks. In the second phase, we develop a novel recursive graph decomposition approach to address this problem. We merge the leaves of the tree that have similar risks to form new leaves that form the final output. We apply Risk-stratify on a cohort of patients (with no history of cardiovascular disease) from UK Biobank and assess their risk for cardiovascular disease. Risk-stratify significantly improves risk-stratification, i.e., a lower fraction of the groups have over/under estimated risks (measured in terms of false discovery rate; 33% reduction) in comparison to state-of-the-art methods for cardiovascular prediction (Random forests, Cox model, etc.). We find that the Cox model significantly over estimates the risk of 21,621 patients out of 216,211 patients. Risk-stratify can accurately categorize 2,987 of these 21,621 patients as low-risk individuals.
Tasks
Published	2018-11-02
URL	http://arxiv.org/abs/1811.00753v1
PDF	http://arxiv.org/pdf/1811.00753v1.pdf
PWC	https://paperswithcode.com/paper/risk-stratify-confident-stratification-of
Repo
Framework

Enriching Frame Representations with Distributionally Induced Senses


Title	Enriching Frame Representations with Distributionally Induced Senses
Authors	Stefano Faralli, Alexander Panchenko, Chris Biemann, Simone Paolo Ponzetto
Abstract	We introduce a new lexical resource that enriches the Framester knowledge graph, which links Framnet, WordNet, VerbNet and other resources, with semantic features from text corpora. These features are extracted from distributionally induced sense inventories and subsequently linked to the manually-constructed frame representations to boost the performance of frame disambiguation in context. Since Framester is a frame-based knowledge graph, which enables full-fledged OWL querying and reasoning, our resource paves the way for the development of novel, deeper semantic-aware applications that could benefit from the combination of knowledge from text and complex symbolic representations of events and participants. Together with the resource we also provide the software we developed for the evaluation in the task of Word Frame Disambiguation (WFD).
Tasks
Published	2018-03-15
URL	http://arxiv.org/abs/1803.05829v1
PDF	http://arxiv.org/pdf/1803.05829v1.pdf
PWC	https://paperswithcode.com/paper/enriching-frame-representations-with
Repo
Framework

Icing on the Cake: An Easy and Quick Post-Learnig Method You Can Try After Deep Learning


Title	Icing on the Cake: An Easy and Quick Post-Learnig Method You Can Try After Deep Learning
Authors	Tomohiko Konno, Michiaki Iwazume
Abstract	We found an easy and quick post-learning method named “Icing on the Cake” to enhance a classification performance in deep learning. The method is that we train only the final classifier again after an ordinary training is done.
Tasks
Published	2018-07-17
URL	http://arxiv.org/abs/1807.06540v1
PDF	http://arxiv.org/pdf/1807.06540v1.pdf
PWC	https://paperswithcode.com/paper/icing-on-the-cake-an-easy-and-quick-post
Repo
Framework

Human-aided Multi-Entity Bayesian Networks Learning from Relational Data


Title	Human-aided Multi-Entity Bayesian Networks Learning from Relational Data
Authors	Cheol Young Park, Kathryn Blackmond Laskey
Abstract	An Artificial Intelligence (AI) system is an autonomous system which emulates human mental and physical activities such as Observe, Orient, Decide, and Act, called the OODA process. An AI system performing the OODA process requires a semantically rich representation to handle a complex real world situation and ability to reason under uncertainty about the situation. Multi-Entity Bayesian Networks (MEBNs) combines First-Order Logic with Bayesian Networks for representing and reasoning about uncertainty in complex, knowledge-rich domains. MEBN goes beyond standard Bayesian networks to enable reasoning about an unknown number of entities interacting with each other in various types of relationships, a key requirement for the OODA process of an AI system. MEBN models have heretofore been constructed manually by a domain expert. However, manual MEBN modeling is labor-intensive and insufficiently agile. To address these problems, an efficient method is needed for MEBN modeling. One of the methods is to use machine learning to learn a MEBN model in whole or in part from data. In the era of Big Data, data-rich environments, characterized by uncertainty and complexity, have become ubiquitous. The larger the data sample is, the more accurate the results of the machine learning approach can be. Therefore, machine learning has potential to improve the quality of MEBN models as well as the effectiveness for MEBN modeling. In this research, we study a MEBN learning framework to develop a MEBN model from a combination of domain expert’s knowledge and data. To evaluate the MEBN learning framework, we conduct an experiment to compare the MEBN learning framework and the existing manual MEBN modeling in terms of development efficiency.
Tasks
Published	2018-06-06
URL	http://arxiv.org/abs/1806.02421v1
PDF	http://arxiv.org/pdf/1806.02421v1.pdf
PWC	https://paperswithcode.com/paper/human-aided-multi-entity-bayesian-networks
Repo
Framework

Generalize Symbolic Knowledge With Neural Rule Engine


Title	Generalize Symbolic Knowledge With Neural Rule Engine
Authors	Shen Li, Hengru Xu, Zhengdong Lu
Abstract	As neural networks have dominated the state-of-the-art results in a wide range of NLP tasks, it attracts considerable attention to improve the performance of neural models by integrating symbolic knowledge. Different from existing works, this paper investigates the combination of these two powerful paradigms from the knowledge-driven side. We propose Neural Rule Engine (NRE), which can learn knowledge explicitly from logic rules and then generalize them implicitly with neural networks. NRE is implemented with neural module networks in which each module represents an action of a logic rule. The experiments show that NRE could greatly improve the generalization abilities of logic rules with a significant increase in recall. Meanwhile, the precision is still maintained at a high level.
Tasks
Published	2018-08-30
URL	https://arxiv.org/abs/1808.10326v3
PDF	https://arxiv.org/pdf/1808.10326v3.pdf
PWC	https://paperswithcode.com/paper/generalize-symbolic-knowledge-with-neural
Repo
Framework