Improving Textless Spoken Language Understanding with Discrete Units as Intermediate Target

no code implementations • 29 May 2023 • Guan-Wei Wu, Guan-Ting Lin, Shang-Wen Li, Hung-Yi Lee

However, the absence of intermediate targets and training guidance for textless SLU often results in suboptimal performance.

Automatic Speech Recognition Automatic Speech Recognition (ASR) +3