Daily Papers — 2026-10-01

1 papers on audio, speech, music, and acoustics.

1. Q-SPT: Learnable Query-Based Compression for Low-Frame-Rate Speech Tokenization

Authors: Jeeyoung Yun, Seohwan Yun, Sungwoong Kim

Categories: cs.SD, cs.CL, eess.AS Score: 4.84/10 (Obj:7 Id:4 Ind:6 Comp:6 Eff:5 Nov:4 Rep:3)

• Strength: 文本自回归损失把 6.25 Hz 语义流的 WER 从 4.97 降到 4.29，下游 ASR WER 降到 5.88，略优于 FlexiCodec 的 6.18。
• Weakness: 缺同为 6.25 Hz 且开源的 TaDiCodec 对比，也未引用已提出 query 压缩加 AR 损失的 ALMTokenizer，全文无方差与主观听测，TTS WER 还输 FlexiCodec 1.17。
• Full review: Claude Code 全文七公理审稿

Neural speech codecs increasingly serve as tokenizers for speech language models (SLMs). Lowering the frame rate reduces the computational and memory costs of SLMs, but makes it difficult to preserve both linguistic information and acoustic detail. Existing approaches rely on rule-based compression: average pooling can discard linguistic information, whereas similarity-based merging uses a fixed threshold on adjacent-frame similarity and applies the resulting boundaries to the acoustic stream. We propose Q-SPT, a low-frame-rate dual-stream speech tokenizer with separate, context-aware, learnable query-based compressors specialized for semantic and acoustic representations. In particular, queries at a fixed rate independently attend to the semantic and acoustic streams as separate key-value sources, enabling stream-specific, context-aware aggregation through two separately learned compressors. In addition, an autoregressive text loss explicitly supervises the semantic compressor to preserve linguistic information. Experimental results show that Q-SPT achieves the best reconstruction among the evaluated codecs at the same frame rate. In downstream SLMs, it yields the best speech recognition accuracy and text-to-speech perceptual quality with competitive intelligibility.