Alternating Weak Triphone/BPE Alignment Supervision from Hybrid Model Improves End-to-End ASR

Autor:	Jiang, Jintao, Gao, Yingbo, Zeineldeen, Mohammad, Tuske, Zoltan
Rok vydání:	2024
Předmět:	Computer Science - Computation and Language Computer Science - Sound Electrical Engineering and Systems Science - Audio and Speech Processing
Druh dokumentu:	Working Paper
Popis:	In this paper, alternating weak triphone/BPE alignment supervision is proposed to improve end-to-end model training. Towards this end, triphone and BPE alignments are extracted using a pre-existing hybrid ASR system. Then, regularization effect is obtained by cross-entropy based intermediate auxiliary losses computed on such alignments at a mid-layer representation of the encoder for triphone alignments and at the encoder for BPE alignments. Weak supervision is achieved through strong label smoothing with parameter of 0.5. Experimental results on TED-LIUM 2 indicate that either triphone or BPE alignment based weak supervision improves ASR performance over standard CTC auxiliary loss. Moreover, their combination lowers the word error rate further. We also investigate the alternation of the two auxiliary tasks during model training, and additional performance gain is observed. Overall, the proposed techniques result in over 10% relative error rate reduction over a CTC-regularized baseline system. Comment: 5 pages, 1 figure, 3 tables
Databáze:	arXiv
Externí odkaz:	http://arxiv.org/abs/2402.15594 Zobrazit plný text záznamu View this record from Arxiv