Search papers, labs, and topics across Lattice.
This paper introduces ABX-Accent, a benchmark for evaluating unsupervised speech representation learning across ten different English accents, addressing the challenge of adapting models to out-of-domain speech. The authors demonstrate that their approach, which employs adaptive domain normalization to fine-tune a pretrained Contrastive Predictive Coding model, significantly enhances performance on accented speech tasks. Notably, their method achieves a 23.6% relative improvement in across-speaker ABX scores compared to non-adapted models, highlighting its effectiveness in domain adaptation for speech recognition.
Unsupervised speech models can be dramatically improved for accented speech, achieving a 23.6% boost in performance with minimal adaptation.
Representation learning has attracted great atten- tion and managed to reach good performances as a pretraining method for downstream tasks or as a first step towards unsu- pervised speech modeling. Yet, little is known about how such methods deal with out-of-domain speech and how could they be adapted in a few shot to new domains. This is important especially for accented speech where one observes a long tail of accents that diverge from the standard ones. We introduce ABX- Accent, a benchmark based on the AESRC dataset that features 10 different accents of English. It includes a small (<10 hours) unlabelled training set in each of the accents and adaptations of the Zero Resources Challenge ABX evaluation metrics to each of the accents. We illustrate this benchmark with a baseline model that uses adaptive domain normalization to fine tune a pretrained Contrastive Predictive Coding model on the accents. This method is first developed on LibriSpeech using a male/female split. When applied to the new benchmark, the proposed method yields a relative improvement of 23.6% on across-speaker ABX scores on average compared to non adapted models. The data and metrics will be open sourced upon paper acceptance
