October 15, 2019

2067 words 10 mins read

Paper Group NANR 193

Paper Group NANR 193

Burst Image Deblurring Using Permutation Invariant Convolutional Neural Networks. Automated Paraphrase Lattice Creation for HyTER Machine Translation Evaluation. Speed Reading: Learning to Read ForBackward via Shuttle. Textual Deconvolution Saliency (TDS) : a deep tool box for linguistic analysis. Text Mining for History: first steps on building a …

Burst Image Deblurring Using Permutation Invariant Convolutional Neural Networks

Title Burst Image Deblurring Using Permutation Invariant Convolutional Neural Networks
Authors Miika Aittala, Fredo Durand
Abstract We propose a neural approach for fusing an arbitrary-length burst of photographs suffering from severe camera shake and noise into a sharp and noise-free image. Our novel convolutional architecture has a simultaneous view of all frames in the burst, and by construction treats them in an order-independent manner. This enables it to effectively detect and leverage subtle cues scattered across different frames, while ensuring that each frame gets a full and equal consideration regardless of its position in the sequence. We train the network with richly varied synthetic data consisting of camera shake, realistic noise, and other common imaging defects. The method demonstrates consistent state of the art burst image restoration performance for highly degraded sequences of real-world images, and extracts accurate detail that is not discernible from any of the individual frames in isolation.
Tasks Deblurring, Image Restoration
Published 2018-09-01
URL http://openaccess.thecvf.com/content_ECCV_2018/html/Miika_Aittala_Burst_Image_Deblurring_ECCV_2018_paper.html
PDF http://openaccess.thecvf.com/content_ECCV_2018/papers/Miika_Aittala_Burst_Image_Deblurring_ECCV_2018_paper.pdf
PWC https://paperswithcode.com/paper/burst-image-deblurring-using-permutation
Repo
Framework

Automated Paraphrase Lattice Creation for HyTER Machine Translation Evaluation

Title Automated Paraphrase Lattice Creation for HyTER Machine Translation Evaluation
Authors Marianna Apidianaki, Guillaume Wisniewski, Anne Cocos, Chris Callison-Burch
Abstract We propose a variant of a well-known machine translation (MT) evaluation metric, HyTER (Dreyer and Marcu, 2012), which exploits reference translations enriched with meaning equivalent expressions. The original HyTER metric relied on hand-crafted paraphrase networks which restricted its applicability to new data. We test, for the first time, HyTER with automatically built paraphrase lattices. We show that although the metric obtains good results on small and carefully curated data with both manually and automatically selected substitutes, it achieves medium performance on much larger and noisier datasets, demonstrating the limits of the metric for tuning and evaluation of current MT systems.
Tasks Machine Translation
Published 2018-06-01
URL https://www.aclweb.org/anthology/N18-2077/
PDF https://www.aclweb.org/anthology/N18-2077
PWC https://paperswithcode.com/paper/automated-paraphrase-lattice-creation-for
Repo
Framework

Speed Reading: Learning to Read ForBackward via Shuttle

Title Speed Reading: Learning to Read ForBackward via Shuttle
Authors Tsu-Jui Fu, Wei-Yun Ma
Abstract We present LSTM-Shuttle, which applies human speed reading techniques to natural language processing tasks for accurate and efficient comprehension. In contrast to previous work, LSTM-Shuttle not only reads shuttling forward but also goes back. Shuttling forward enables high efficiency, and going backward gives the model a chance to recover lost information, ensuring better prediction. We evaluate LSTM-Shuttle on sentiment analysis, news classification, and cloze on IMDB, Rotten Tomatoes, AG, and Children{'}s Book Test datasets. We show that LSTM-Shuttle predicts both better and more quickly. To demonstrate how LSTM-Shuttle actually behaves, we also analyze the shuttling operation and present a case study.
Tasks Document Classification, Document Summarization, Machine Translation, Named Entity Recognition, Part-Of-Speech Tagging, Question Answering, Reading Comprehension, Sentiment Analysis
Published 2018-10-01
URL https://www.aclweb.org/anthology/D18-1474/
PDF https://www.aclweb.org/anthology/D18-1474
PWC https://paperswithcode.com/paper/speed-reading-learning-to-read-forbackward
Repo
Framework

Textual Deconvolution Saliency (TDS) : a deep tool box for linguistic analysis

Title Textual Deconvolution Saliency (TDS) : a deep tool box for linguistic analysis
Authors Laurent Vanni, Melanie Ducoffe, Carlos Aguilar, Frederic Precioso, Damon Mayaffre
Abstract In this paper, we propose a new strategy, called Text Deconvolution Saliency (TDS), to visualize linguistic information detected by a CNN for text classification. We extend Deconvolution Networks to text in order to present a new perspective on text analysis to the linguistic community. We empirically demonstrated the efficiency of our Text Deconvolution Saliency on corpora from three different languages: English, French, and Latin. For every tested dataset, our Text Deconvolution Saliency automatically encodes complex linguistic patterns based on co-occurrences and possibly on grammatical and syntax analysis.
Tasks Text Classification
Published 2018-07-01
URL https://www.aclweb.org/anthology/P18-1051/
PDF https://www.aclweb.org/anthology/P18-1051
PWC https://paperswithcode.com/paper/textual-deconvolution-saliency-tds-a-deep
Repo
Framework

Text Mining for History: first steps on building a large dataset

Title Text Mining for History: first steps on building a large dataset
Authors Suemi Higuchi, Cl{'a}udia Freitas, Bruno Cuconato, Alex Rademaker, re
Abstract
Tasks
Published 2018-05-01
URL https://www.aclweb.org/anthology/L18-1593/
PDF https://www.aclweb.org/anthology/L18-1593
PWC https://paperswithcode.com/paper/text-mining-for-history-first-steps-on
Repo
Framework

Effects of Gender Stereotypes on Trust and Likability in Spoken Human-Robot Interaction

Title Effects of Gender Stereotypes on Trust and Likability in Spoken Human-Robot Interaction
Authors Matthias Kraus, Johannes Kraus, Martin Baumann, Wolfgang Minker
Abstract
Tasks
Published 2018-05-01
URL https://www.aclweb.org/anthology/L18-1018/
PDF https://www.aclweb.org/anthology/L18-1018
PWC https://paperswithcode.com/paper/effects-of-gender-stereotypes-on-trust-and
Repo
Framework

Looking for Structure in Lexical and Acoustic-Prosodic Entrainment Behaviors

Title Looking for Structure in Lexical and Acoustic-Prosodic Entrainment Behaviors
Authors Andreas Weise, Rivka Levitan
Abstract Entrainment has been shown to occur for various linguistic features individually. Motivated by cognitive theories regarding linguistic entrainment, we analyze speakers{'} overall entrainment behaviors and search for an underlying structure. We consider various measures of both acoustic-prosodic and lexical entrainment, measuring the latter with a novel application of two previously introduced methods in addition to a standard high-frequency word measure. We present a negative result of our search, finding no meaningful correlations, clusters, or principal components in various entrainment measures, and discuss practical and theoretical implications.
Tasks
Published 2018-06-01
URL https://www.aclweb.org/anthology/N18-2048/
PDF https://www.aclweb.org/anthology/N18-2048
PWC https://paperswithcode.com/paper/looking-for-structure-in-lexical-and-acoustic
Repo
Framework

Interpretable Word Embedding Contextualization

Title Interpretable Word Embedding Contextualization
Authors Kyoung-Rok Jang, Sung-Hyon Myaeng, Sang-Bum Kim
Abstract In this paper, we propose a method of calibrating a word embedding, so that the semantic it conveys becomes more relevant to the context. Our method is novel because the output shows clearly which senses that were originally presented in a target word embedding become stronger or weaker. This is possible by utilizing the technique of using sparse coding to recover senses that comprises a word embedding.
Tasks Word Embeddings
Published 2018-11-01
URL https://www.aclweb.org/anthology/W18-5442/
PDF https://www.aclweb.org/anthology/W18-5442
PWC https://paperswithcode.com/paper/interpretable-word-embedding
Repo
Framework

LEARNING SEMANTIC WORD RESPRESENTATIONS VIA TENSOR FACTORIZATION

Title LEARNING SEMANTIC WORD RESPRESENTATIONS VIA TENSOR FACTORIZATION
Authors Eric Bailey, Charles Meyer, Shuchin Aeron
Abstract Many state-of-the-art word embedding techniques involve factorization of a cooccurrence based matrix. We aim to extend this approach by studying word embedding techniques that involve factorization of co-occurrence based tensors (N- way arrays). We present two new word embedding techniques based on tensor factorization and show that they outperform common methods on several semantic NLP tasks when given the same data. To train one of the embeddings, we present a new joint tensor factorization problem and an approach for solving it. Furthermore, we modify the performance metrics for the Outlier Detection Camacho- Collados & Navigli (2016) task to measure the quality of higher-order relationships that a word embedding captures. Our tensor-based methods significantly outperform existing methods at this task when using our new metric. Finally, we demonstrate that vectors in our embeddings can be composed multiplicatively to create different vector representations for each meaning of a polysemous word. We show that this property stems from the higher order information that the vectors contain, and thus is unique to our tensor based embeddings.
Tasks Outlier Detection
Published 2018-01-01
URL https://openreview.net/forum?id=B1kIr-WRb
PDF https://openreview.net/pdf?id=B1kIr-WRb
PWC https://paperswithcode.com/paper/learning-semantic-word-respresentations-via
Repo
Framework

Natural Language Inference with Definition Embedding Considering Context On the Fly

Title Natural Language Inference with Definition Embedding Considering Context On the Fly
Authors Kosuke Nishida, Kyosuke Nishida, Hisako Asano, Junji Tomita
Abstract Natural language inference (NLI) is one of the most important tasks in NLP. In this study, we propose a novel method using word dictionaries, which are pairs of a word and its definition, as external knowledge. Our neural definition embedding mechanism encodes input sentences with the definitions of each word of the sentences on the fly. It can encode the definition of words considering the context of input sentences by using an attention mechanism. We evaluated our method using WordNet as a dictionary and confirmed that our method performed better than baseline models when using the full or a subset of 100d GloVe as word embeddings.
Tasks Domain Adaptation, Information Retrieval, Natural Language Inference, Question Answering, Representation Learning, Word Embeddings
Published 2018-07-01
URL https://www.aclweb.org/anthology/W18-3007/
PDF https://www.aclweb.org/anthology/W18-3007
PWC https://paperswithcode.com/paper/natural-language-inference-with-definition
Repo
Framework

Adaptive Negative Curvature Descent with Applications in Non-convex Optimization

Title Adaptive Negative Curvature Descent with Applications in Non-convex Optimization
Authors Mingrui Liu, Zhe Li, Xiaoyu Wang, Jinfeng Yi, Tianbao Yang
Abstract Negative curvature descent (NCD) method has been utilized to design deterministic or stochastic algorithms for non-convex optimization aiming at finding second-order stationary points or local minima. In existing studies, NCD needs to approximate the smallest eigen-value of the Hessian matrix with a sufficient precision (e.g., $\epsilon_2\ll 1$) in order to achieve a sufficiently accurate second-order stationary solution (i.e., $\lambda_{\min}(\nabla^2 f(\x))\geq -\epsilon_2)$. One issue with this approach is that the target precision $\epsilon_2$ is usually set to be very small in order to find a high quality solution, which increases the complexity for computing a negative curvature. To address this issue, we propose an adaptive NCD to allow for an adaptive error dependent on the current gradient’s magnitude in approximating the smallest eigen-value of the Hessian, and to encourage competition between a noisy NCD step and gradient descent step. We consider the applications of the proposed adaptive NCD for both deterministic and stochastic non-convex optimization, and demonstrate that it can help reduce the the overall complexity in computing the negative curvatures during the course of optimization without sacrificing the iteration complexity.
Tasks
Published 2018-12-01
URL http://papers.nips.cc/paper/7734-adaptive-negative-curvature-descent-with-applications-in-non-convex-optimization
PDF http://papers.nips.cc/paper/7734-adaptive-negative-curvature-descent-with-applications-in-non-convex-optimization.pdf
PWC https://paperswithcode.com/paper/adaptive-negative-curvature-descent-with
Repo
Framework

A farewell to arms: Non-verbal communication for non-humanoid robots

Title A farewell to arms: Non-verbal communication for non-humanoid robots
Authors Aaron G. Cass, Kristina Striegnitz, Nick Webb
Abstract Human-robot interactions situated in a dynamic environment create a unique mix of challenges for conversational systems. We argue that, on the one hand, NLG can contribute to addressing these challenges and that, on the other hand, they pose interesting research problems for NLG. To illustrate our position we describe our research on non-humanoid robots using non-verbal signals to support communication.
Tasks Text Generation
Published 2018-11-01
URL https://www.aclweb.org/anthology/W18-6905/
PDF https://www.aclweb.org/anthology/W18-6905
PWC https://paperswithcode.com/paper/a-farewell-to-arms-non-verbal-communication
Repo
Framework

Neural Tree Transducers for Tree to Tree Learning

Title Neural Tree Transducers for Tree to Tree Learning
Authors João Sedoc, Dean Foster, Lyle Ungar
Abstract We introduce a novel approach to tree-to-tree learning, the neural tree transducer (NTT), a top-down depth first context-sensitive tree decoder, which is paired with recursive neural encoders. Our method works purely on tree-to-tree manipulations rather than sequence-to-tree or tree-to-sequence and is able to encode and decode multiple depth trees. We compare our method to sequence-to-sequence models applied to serializations of the trees and show that our method outperforms previous methods for tree-to-tree transduction.
Tasks
Published 2018-01-01
URL https://openreview.net/forum?id=rJBwoM-Cb
PDF https://openreview.net/pdf?id=rJBwoM-Cb
PWC https://paperswithcode.com/paper/neural-tree-transducers-for-tree-to-tree
Repo
Framework

A Comparison of Two Paraphrase Models for Taxonomy Augmentation

Title A Comparison of Two Paraphrase Models for Taxonomy Augmentation
Authors Vassilis Plachouras, Fabio Petroni, Timothy Nugent, Jochen L. Leidner
Abstract Taxonomies are often used to look up the concepts they contain in text documents (for instance, to classify a document). The more comprehensive the taxonomy, the higher recall the application has that uses the taxonomy. In this paper, we explore automatic taxonomy augmentation with paraphrases. We compare two state-of-the-art paraphrase models based on Moses, a statistical Machine Translation system, and a sequence-to-sequence neural network, trained on a paraphrase datasets with respect to their abilities to add novel nodes to an existing taxonomy from the risk domain. We conduct component-based and task-based evaluations. Our results show that paraphrasing is a viable method to enrich a taxonomy with more terms, and that Moses consistently outperforms the sequence-to-sequence neural model. To the best of our knowledge, this is the first approach to augment taxonomies with paraphrases.
Tasks Document Classification, Machine Translation, Paraphrase Generation
Published 2018-06-01
URL https://www.aclweb.org/anthology/N18-2051/
PDF https://www.aclweb.org/anthology/N18-2051
PWC https://paperswithcode.com/paper/a-comparison-of-two-paraphrase-models-for
Repo
Framework

A Laypeople Study on Terminology Identification across Domains and Task Definitions

Title A Laypeople Study on Terminology Identification across Domains and Task Definitions
Authors Anna H{"a}tty, Sabine Schulte im Walde
Abstract This paper introduces a new dataset of term annotation. Given that even experts vary significantly in their understanding of termhood, and that term identification is mostly performed as a binary task, we offer a novel perspective to explore the common, natural understanding of what constitutes a term: Laypeople annotate single-word and multi-word terms, across four domains and across four task definitions. Analyses based on inter-annotator agreement offer insights into differences in term specificity, term granularity and subtermhood.
Tasks Information Retrieval
Published 2018-06-01
URL https://www.aclweb.org/anthology/N18-2052/
PDF https://www.aclweb.org/anthology/N18-2052
PWC https://paperswithcode.com/paper/a-laypeople-study-on-terminology
Repo
Framework
comments powered by Disqus