Learning Contextualized Representation On Discrete Space Via Hierarchical Product Quantization

AuthorHYUNG YONG KIM, BYEONG-YEOL KIM, YUNKYU Lim, JIHWAN PARK, JINSEOK PARK, YOUSHIN LIM, SEUNG WOO YU, HANBIN LEE

Abstract

Self-supervised learning has recently demonstrated significant success in various speech processing applications. Recent studies report that pre-training with contextualized continuous targets plays a crucial role in fine-tuning for better speech downstream tasks. However, unlike the continuous targets, it is challenging to produce contextualized targets on discrete space due to unstable training.

  • ConferenceICASSP
  • TopicSpeech Recognition
  • Year Published2024

Abstract

Self-supervised learning has recently demonstrated significant success in various speech processing applications. Recent studies report that pre-training with contextualized continuous targets plays a crucial role in fine-tuning for better speech downstream tasks. However, unlike the continuous targets, it is challenging to produce contextualized targets on discrete space due to unstable training.