Timers and Such: A Practical Benchmark for Spoken Language Understanding with Numbers (2021-04-04T00:00:00.000000Z)

TL;DR

The gap in existing spoken language understanding datasets that Timers and Such fills is described, the design and creation of the dataset are described, and experiments with a number of ASR-based and end-to-end baseline models are experiments.

Abstract

This paper introduces Timers and Such, a new open source dataset of spoken English commands for common voice control use cases involving numbers. We describe the gap in existing spoken language understanding datasets that Timers and Such fills, the design and creation of the dataset, and experiments with a number of ASR-based and end-to-end baseline models, the code for which has been made available as part of the SpeechBrain toolkit.

Authors

M. Ravanelli

17 papers

Loren Lugosch

4 papers

Piyush Papreja

1 papers

TL;DR

Abstract

Authors

References43 items

SPLAT: Speech-Language Joint Pre-Training for Spoken Language Understanding

SpeechBrain: A General-Purpose Speech Toolkit

A Streaming End-to-End Framework For Spoken Language Understanding

Integration of Pre-Trained Networks with Continuous Token Interface for End-to-End Spoken Language Understanding

Leveraging Acoustic and Linguistic Embeddings from Pretrained Speech and Language Models for Intent Classification

SLURP: A Spoken Language Understanding Resource Package

A Low Latency ASR-Free End to End Spoken Language Understanding System

Rethinking Evaluation in ASR: Are Our Models Robust Enough?

Viewmaker Networks: Learning Views for Unsupervised Representation Learning

Semi-Supervised Speech-Language Joint Pre-Training for Spoken Language Understanding

wav2vec 2.0: A Framework for Self-Supervised Learning of Speech Representations

Leveraging Unpaired Text Data for Training End-To-End Speech-to-Intent Systems

Large-Scale Unsupervised Pre-Training for End-to-End Spoken Language Understanding

Common Voice: A Massively-Multilingual Speech Corpus

Using Speech Synthesis to Train End-To-End Spoken Language Understanding Models

Reducing malicious use of synthetic media research: Considerations and potential release practices for machine learning

Speech Model Pre-training for End-to-End Spoken Language Understanding

Learning Problem-agnostic Speech Representations from Multiple Self-supervised Tasks

Spoken Language Understanding on the Edge

From Audio to Semantics: Approaches to End-to-End Spoken Language Understanding

SentencePiece: A simple and language independent subword tokenizer and detokenizer for Neural Text Processing

Transfer Learning from Speaker Verification to Multispeaker Text-To-Speech Synthesis

Speech Commands: A Dataset for Limited-Vocabulary Speech Recognition

Towards End-to-end Spoken Language Understanding

An Analysis of Incorporating an External Language Model into a Sequence-to-Sequence Model

VoiceLoop: Voice Fitting and Synthesis via a Phonological Loop

SUPERSEDED - CSTR VCTK Corpus: English Multi-speaker Corpus for CSTR Voice Cloning Toolkit

Joint CTC-attention based end-to-end speech recognition using multi-task learning

Listen, attend and spell: A neural network for large vocabulary conversational speech recognition

Attention-Based Models for Speech Recognition

Librispeech: An ASR corpus based on public domain audio books

Neural Machine Translation by Jointly Learning to Align and Translate

Sequence Transduction with Recurrent Neural Networks

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

Language Understanding

Connectionist Temporal Classiﬁcation: Labelling Unsegmented Sequence Data with Recurrent Neural Networks

Jakobovski/free-spoken-digit-dataset: v1.0.8

Tidigits ldc93s10

Master thesis: Real-time voice cloning

Speech and Signal Processing (ICASSP)

DARPA TIMIT Acoustic Phonetic Continuous Speech Corpus CDROM,

You will see a prompt displayed on the screen. Click on the red button to record the prompt

where it can be easily downloaded either through the browser or using a

Field of Study

Journal Information

Name

Volume