Showing eess.ASShow all
2 papers · 1 filter
eess.AS2026
Audio Tokens as a Budgeted Resource: Marginal-Utility Allocation for Scalable Audio Representations
Mingyu Zhao, Jinchao Zhang, Zhiyong Wu
Discrete audio tokens are widely used as a representation interface, yet fixed-depth RVQ tokenizers allocate equal capacity to every frame despite varying refinement value. We intr…
eess.AS2026
BAMU: Bitstream-Aware Marginal-Utility Allocation for Frozen Pretrained Neural Speech Codecs
Mingyu Zhao, Zijian Lin, Yutang Feng +6
Pretrained neural speech codecs typically use a fixed residual vector quantization (RVQ) depth for all frames, ignoring temporal variation in quantization difficulty. We propose BA…