Electrical Engineering and Systems Science > Audio and Speech Processing
[Submitted on 19 Jun 2020 (this version), latest version 5 Nov 2020 (v2)]
Title:Efficient Active Learning for Automatic Speech Recognition via Augmented Consistency Regularization
View PDFAbstract:The cost of labeling transcriptions for large speech corpora becomes a bottleneck to maximally enjoy the potential capacity of deep neural network-based automatic speech recognition (ASR) models. Therefore, in this paper, we present a new training scheme that minimizes the labeling cost by adopting the concepts of semi-supervised learning (SSL) and active learning (AL) approaches and making a synergy from them. While AL studies only focus on selecting minimized the number of samples to be labeled with a criterion and taking advantage of such samples, we show that the training efficiency can be further improved by utilizing the unlabeled samples by sophisticatedly designing unsupervised loss that complements the unwanted behavior of supervised loss effectively. Our unsupervised loss is built on Consistency-Regularization (CR) approach, and we propose appropriate augmentation techniques to adopt CR in ASR field successfully. From the qualitative and quantitative experiments on the real-world dataset from deployed end-user voice assistant services, we show that the proposed methods can handle a large number of unlabeled speech data to achieve competitive model performance, with a sustainable amount of human labeling cost.
Submission history
From: Heesu Kim [view email][v1] Fri, 19 Jun 2020 08:54:46 UTC (166 KB)
[v2] Thu, 5 Nov 2020 14:41:47 UTC (230 KB)
Current browse context:
eess.AS
References & Citations
Bibliographic and Citation Tools
Bibliographic Explorer (What is the Explorer?)
Connected Papers (What is Connected Papers?)
Litmaps (What is Litmaps?)
scite Smart Citations (What are Smart Citations?)
Code, Data and Media Associated with this Article
alphaXiv (What is alphaXiv?)
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub (What is DagsHub?)
Gotit.pub (What is GotitPub?)
Hugging Face (What is Huggingface?)
Papers with Code (What is Papers with Code?)
ScienceCast (What is ScienceCast?)
Demos
Recommenders and Search Tools
Influence Flower (What are Influence Flowers?)
CORE Recommender (What is CORE?)
arXivLabs: experimental projects with community collaborators
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.