AI Paper Atlas

Research map / Video / Video-language understanding

Say More with Less: Variable-Frame-Rate Speech Tokenization via Adaptive Clustering and Implicit Duration Coding

Rui-Chen Zheng, Wenrui Liu, Hui-Peng Du, Qinglin Zhang, Chong Deng, Qian Chen, Wen Wang, Yang Ai, Zhen-Hua Ling

AAAI 2026 · Video › Video-language understanding · approach: Video LLMs with fine-grained understanding · 4 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use variable-frame-rate clustering (tau 0.7, Smax 4, 30.95 Hz) instead of fixed 40 HzWavTokenizer · LibriTTS test-clean reconstructionUTMOS: 3.6107 -> 3.8949Improved
variable rate (tau 0.7, Smax 4) vs fixed 40 HzWavTokenizer · LibriTTS test-clean reconstructionPESQ / STOI: PESQ 1.7075 -> 1.7095; STOI 0.8652 -> 0.8601No clear effect
variable rate (tau 0.8) vs 75 Hz fixedWavTokenizer · LibriTTS test-clean reconstructionUTMOS: 4.0247 -> 4.0000No clear effect
raise similarity threshold tau from 0.6 to 0.8 (ablation)VARSTok · LibriTTS test-clean reconstructionUTMOS: 3.83 -> 4.00 (frame rate rises 26.29 -> 36.81 Hz)Improved
increase max span Smax from 2 to 8 (ablation)VARSTok · LibriTTS test-clean reconstructionUTMOS / PESQ: UTMOS 4.0379 -> 3.6466; PESQ 2.0694 -> 1.4532 (46.50 -> 22.38 Hz)Hurt
variable-rate tokens vs fixed 40 Hz for semantic probingWavTokenizer · ARCH AudioMNISTaccuracy: 0.4597 -> 0.6111Improved
variable-rate tokens vs fixed 40 Hz for semantic probingWavTokenizer · ARCH RAVDESSaccuracy: 0.2847 -> 0.2951 (tau 0.7); 0.2639 (tau 0.8)Mixed
use VARSTok (tau 0.8) tokens in AR TTS vs 40 Hz WavTokenizerdecoder-only AR TTS · UniCATS zero-shot TTSWER (%): 7.481 -> 6.787Improved
use VARSTok (tau 0.8) tokens in AR TTSdecoder-only AR TTS · UniCATS zero-shot TTSUTMOS: 3.920 -> 4.246Improved
lower token rate (tau 0.7) in TTSdecoder-only AR TTS · UniCATS zero-shot TTSspeaker similarity: 0.918 -> 0.895Hurt
lower tau to 0.6 (26.29 Hz) in TTS (ablation)VARSTok · UniCATS zero-shot TTSWER (%): 7.294 -> 9.393Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy