Publication Details

Ask2Mask: Guided Data Selection for Masked Speech Modeling

BASKAR, M.; ROSENBERG, A.; RAMABHADRAN, B.; ZHANG, Y.; MORENO, P. Ask2Mask: Guided Data Selection for Masked Speech Modeling. IEEE J-STSP, 2022, vol. 16, no. 6, p. 1357-1366. ISSN: 1932-4553.
Czech title
Ask2Mask: Řízený výběr dat pro modelování uměle maskované řeči
Type
journal article
Language
English
Authors
Baskar Murali Karthick, Ing., Ph.D.
Rosenberg Andrew
Ramabhadran Bhuvana
Zhang Yu
Moreno Pedro
URL
Keywords

Guided Data Selection, Masked Speech Modeling

Abstract

Masked speech modeling (MSM) methods such as wav2vec2 or w2v-BERT learn
representations over speech frames which are randomlymaskedwithin an utterance.
While thesemethods improve performance of Automatic Speech Recognition (ASR)
systems, they have one major limitation. They treat all unsupervised speech
samples with equal weight, which hinders learning as not all samples have
relevant information to learn meaningful representations. In this work, we
address this limitation. We propose ask2mask (ATM), a novel approach to focus on
specific samples during MSM pre-training. ATM employs an external ASR model or
scorer to weight unsupervised input samples in two different ways: 1)
A fine-grained data selection is performed by masking over the highly confident
input frames as chosen by the scorer. This allows themodel to learnmeaningful
representations. 2) ATM is further extended to focus at utterance-level by
weighting the final MSM loss with the utterance-level confidence score. We
conduct fine-tuning experiments on two well-benchmarked corpora: LibriSpeech
(matching the pre-training data) and Commonvoice, TED-LIUM, AMI and CHiME-6 (not
matching the pre-training data). The results substantiate the efficacy of ATM on
significantly improving the recognition performance under mismatched conditions
(up to 11.6% relative over published results and upto 4.46% relative over our
internal baseline) while still yielding modest improvements under matched
conditions.

Published
2022
Pages
1357–1366
Journal
IEEE J-STSP, vol. 16, no. 6, ISSN 1932-4553
DOI
UT WoS
000870301500019
EID Scopus
BibTeX
@article{BUT182529,
  author="Murali Karthick {Baskar} and Andrew {Rosenberg} and Bhuvana {Ramabhadran} and Yu {Zhang} and Pedro {Moreno}",
  title="Ask2Mask: Guided Data Selection for Masked Speech Modeling",
  journal="IEEE J-STSP",
  year="2022",
  volume="16",
  number="6",
  pages="1357--1366",
  doi="10.1109/JSTSP.2022.3186162",
  issn="1932-4553",
  url="https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=9806175"
}
Files
Back to top