Spoken SQuAD: A… (2018-04-01T00:00:00.000000Z)

TL;DR

On the new listening comprehension task, it is found that speech recognition errors have catastrophic impact on machine comprehension, and several approaches are proposed to mitigate the impact.

Abstract

Reading comprehension has been widely studied. One of the most representative reading comprehension tasks is Stanford Question Answering Dataset (SQuAD), on which machine is already comparable with human. On the other hand, accessing large collections of multimedia or spoken content is much more difficult and time-consuming than plain text content for humans. It's therefore highly attractive to develop machines which can automatically understand spoken content. In this paper, we propose a new listening comprehension task - Spoken SQuAD. On the new task, we found that speech recognition errors have catastrophic impact on machine comprehension, and several approaches are proposed to mitigate the impact.

Authors

Chia-Hsuan Lee

3 papers

Hung-yi Lee

13 papers

Szu-Lin Wu

1 papers

TL;DR

Abstract

Authors

References36 items

FusionNet: Fusing via Fully-Aware Attention with Application to Machine Comprehension

Improving the understanding of spoken referring expressions through syntactic-semantic and contextual-phonetic error-correction

Gated Self-Matching Networks for Reading Comprehension and Question Answering

ASR Error Management for Improving Spoken Language Understanding

Reinforced Mnemonic Reader for Machine Comprehension

TL;DR

Abstract

Authors

References36 items

FusionNet: Fusing via Fully-Aware Attention with Application to Machine Comprehension

Improving the understanding of spoken referring expressions through syntactic-semantic and contextual-phonetic error-correction

Gated Self-Matching Networks for Reading Comprehension and Question Answering

ASR Error Management for Improving Spoken Language Understanding

Reinforced Mnemonic Reader for Machine Comprehension

Reading Wikipedia to Answer Open-Domain Questions

Abstractive headline generation for spoken content by attentive recurrent neural networks with ASR error modeling

Bidirectional Attention Flow for Machine Comprehension

Phoneme Embedding and its Application to Speech Driven Talking Avatar Synthesis

Multi-Domain Joint Semantic Frame Parsing Using Bi-Directional RNN-LSTM

Acoustic Word Embeddings for ASR Error Detection

Hierarchical attention model for improved machine comprehension of spoken content

Towards Machine Comprehension of Spoken Content: Initial TOEFL Listening Comprehension Test by Machine

Sequential Convolutional Neural Networks for Slot Filling in Spoken Language Understanding

SQuAD: 100,000+ Questions for Machine Comprehension of Text

Combining Continuous Word Representation and Prosodic Features for ASR Error Prediction

Character-level Convolutional Networks for Text Classification

Character-Aware Neural Language Models

Using Recurrent Neural Networks for Slot Filling in Spoken Language Understanding

Knowledge-Powered Deep Learning for Word Embedding

Compositional Morphology for Word Representations and Language Modelling

ASR error detection using recurrent neural network language model and complementary ASR

Joint Discriminative Decoding of Words and Semantic Tags for Spoken Language Understanding

Better Word Representations with Recursive Neural Networks for Morphology

ASR error segment localization for spoken recovery strategy

ASR error detection in a conversational spoken language translation system

Sibyl, a factoid question-answering system for spoken documents

Factoid question answering for spoken documents

Overview of QAST 2008

Beyond ASR 1-best: Using word confusion networks in spoken language understanding

Sphinx-4: a flexible open source framework for speech recognition

Improving spoken language understanding using word confusion networks

Learning Knowledge Graphs for Question Answering through Conversational Dialog

Exploring the use of Attention-Based Recurrent Neural Networks For Spoken Language Understanding

Spoken question answering using tree-structured conditional random fields and two-layer random walk

Spoken Question Answering

Field of Study

Venue Information

Name

Type

URL

Alternate Names