TY  - RPRT
TI  - Stride-k Subsampling: Train-Free Audio Token Reduction for Whisper
AU  - Chanhee Cho
AU  - Junhyuk Choi
AU  - Bugeun Kim
PY  - 2026
UR  - https://arxiv.org/abs/2608.30927
ID  - 2608.30927
ER  -