Towards Better Domain Adaptation for Self-Supervised Models: A Case Study of Child ASR

You are here

Top Reasons to Join SPS Today!

1. IEEE Signal Processing Magazine
2. Signal Processing Digital Library*
3. Inside Signal Processing Newsletter
4. SPS Resource Center
5. Career advancement & recognition
6. Discounts on conferences and publications
7. Professional networking
8. Communities for students, young professionals, and women
9. Volunteer opportunities
10. Coming soon! PDH/CEU credits
Click here to learn more.

Towards Better Domain Adaptation for Self-Supervised Models: A Case Study of Child ASR

By: 
Ruchao Fan; Yunzheng Zhu; Jinhan Wang; Abeer Alwan

Recently, self-supervised learning (SSL) from unlabelled speech data has gained increased attention in the automatic speech recognition (ASR) community. Typical SSL methods include autoregressive predictive coding (APC), Wav2vec2.0, and hidden unit BERT (HuBERT). However, SSL models are biased to the pretraining data. When SSL models are finetuned with data from another domain, domain shifting occurs and might cause limited knowledge transfer for downstream tasks. In this paper, we propose a novel framework, domain responsible adaptation and finetuning (DRAFT), to reduce domain shifting in pretrained speech models, and evaluate it for a causal and non-causal transformer. For the causal transformer, an extension of APC (E-APC) is proposed to learn richer information from unlabelled data by using multiple temporally-shifted sequences to perform prediction. For the non-causal transformer, various solutions for using the bidirectional APC (Bi-APC) are investigated. In addition, the DRAFT framework is examined for Wav2vec2.0 and HuBERT methods, which use non-causal transformers as the backbone. The experiments are conducted on child ASR (using the OGI and MyST databases) using SSL models trained with unlabelled adult speech data from Librispeech. The relative WER improvements of up to 19.7% on the two child tasks are observed when compared to the pretrained models without adaptation. With the proposed methods (E-APC and DRAFT), the relative WER improvements are even larger (30% and 19% on the OGI and MyST data, respectively) when compared to the models without using pretraining methods.

Despite impressive advancement in developing automatic speech recognition (ASR) techniques in the last decade, children's ASR remains difficult. Challenges arise, in part, from difficulties in acoustic and language modeling of child speech. Due to different growth patterns of children and motor control issues, child speech has a higher degree of intra-speaker and inter-speaker acoustic variability than adult speech [1]. Additionally, child speech is characterized by significant mispronunciations and disfluencies [2][3]. Another challenge is the lack of large-scale publicly-available child speech databases, and thus child ASR can be treated as a low-resource task [4].

SPS on Twitter

  • New SPS Webinar: On Wednesday, 8 February, join Dr. Roula Nassif for "Decentralized learning over multitask graphs"… https://t.co/GOgHb7vfAv
  • CALL FOR PAPERS: IEEE Signal Processing Magazine welcomes submissions for a Special Issue on Hypercomplex Signal an… https://t.co/UDvjUY2llT
  • New SPS Webinar: On 15 February, join Mr. Wei Liu, Dr. Li Chen and Dr. Wenyi Zhang presenting "Decentralized Federa… https://t.co/em0sQAK4V5
  • New SPS Webinar: On Monday, 13 February, join Dr. Joe (Zhou) Ren when he presents "Human Centric Visual Analysis -… https://t.co/Rc39HpkPKr
  • Help us illustrate the SPS story! In honor of our 75th anniversary, we need your support to capture the people, mem… https://t.co/MnYU9MzIok

SPS Videos


Signal Processing in Home Assistants

 


Multimedia Forensics


Careers in Signal Processing             

 


Under the Radar