Accelerating Hakka Speech Recognition Research and Development Using the Whisper Model

1091

Knowledge Triples