Realizing Petabyte Scale Acoustic Modeling

Large scale machine learning (ML) systems such as the Alexa automatic speech recognition (ASR) system continue to improve with increasing amounts of manually transcribed training data. Instead of scaling manual transcription to impractical levels, we utilize semi-supervised learning (SSL) to learn a...

Full description

Saved in:
Bibliographic Details
Published in:IEEE journal on emerging and selected topics in circuits and systems 2019-06, Vol.9 (2), p.422-432
Main Authors: Parthasarathi, Sree Hari Krishnan, Sivakrishnan, Nitin, Ladkat, Pranav, Strom, Nikko
Format: Article
Language:English
Subjects:
acoustic models
Acoustic noise
Acoustics
Audio data
Automatic speech recognition
Computational modeling
Data models
Distributed databases
Hidden Markov models
large scale semi-supervised learning
Machine learning
Pipelines
Speech recognition
Subsystems
Systems design
Training
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Large scale machine learning (ML) systems such as the Alexa automatic speech recognition (ASR) system continue to improve with increasing amounts of manually transcribed training data. Instead of scaling manual transcription to impractical levels, we utilize semi-supervised learning (SSL) to learn acoustic models (AM) from the vast firehose of untranscribed audio data. Learning an AM from 1 Million hours of audio presents unique ML and system design challenges. We present the design and evaluation of a highly scalable and resource efficient SSL system for AM. Employing the student/teacher learning paradigm, we focus on the student learning subsystem: a scalable and robust data pipeline that generates features and targets from raw audio, and an efficient model pipeline, including the distributed trainer, that builds a student model. Our evaluations show that, even without extensive hyper-parameter tuning, we obtain relative accuracy improvements in the 10 to 20% range, with higher gains in noisier conditions. The end-to-end processing time of this SSL system was 12 days, and several components in this system can trivially scale linearly with more compute resources.
ISSN:2156-3357
2156-3365
DOI:10.1109/JETCAS.2019.2912353