Can I read CTCBERT: Advancing Hidden-Unit Bert with CTC Objectives on EtoBox?

CTCBERT: Advancing Hidden-Unit Bert with CTC Objectives by Ruchao Fan; Yiming Wang; Yashesh Gaur; Jinyu Li is a scholarly article available to read on EtoBox.

What is CTCBERT: Advancing Hidden-Unit Bert with CTC Objectives about?

In this work, we present a simple but effective method, CTCBERT, for advancing hidden-unit BERT (HuBERT). HuBERT applies a frame-level cross-entropy (CE) loss, which is similar to most acoustic model training. However, CTCBERT performs the model training with the Connectionist Temporal Classification (CTC) objective after removing duplicated IDs in each masked region. The idea stems from the observation that there can be significant errors in alignments when using clustered or aligned IDs. CTC learns alignments implicitly, indicating that learning with CTC can be more flexible when misalignment exists. We examine CTCBERT on IDs from HuBERT Iter1, HuBERT Iter2, and PBERT. The CTC training brings consistent improvements compared to the CE training. Furthermore, when loading blank-related parameters during finetuning, slight improvements are observed. Evaluated on the Librispeech 960-100h setting, the relative WER improvements of CTCBERT are 2%-11% over HuBERT and PERT on test-other data.

Author

Ruchao Fan; Yiming Wang; Yashesh Gaur; Jinyu Li

Publisher

IEEE

Published

2023

Language

EN