반응형
[Paper 리뷰] DualCodec: A Low-Frame-Rate, Semantically-Enhanced Neural Audio Codec for Speech Generation
DualCodec: A Low-Frame-Rate, Semantically-Enhanced Neural Audio Codec for Speech GenerationNeural audio codec은 frame rate와 audio quality 간의 trade-off를 가짐DualCodecSelf-Supervised Learning representation과 waveform representation을 integrateFirst-layer codec의 semantic information을 향상하고 low frame rate에서 동작논문 (INTERSPEECH 2025) : Paper Link1. IntroductionNeural audio codec은 audio signal을 discrete code..
Paper/Neural Codec
2025. 7. 5. 07:35
반응형