TY  - RPRT
TI  - Adaptive Two-Stage Visual Token Pruning for Efficient Inference in Video-Language Models
AU  - Paribesh Regmi
AU  - Qingshuang Chen
AU  - Chi Zhang
AU  - Heba Aly
AU  - Yelin Kim
AU  - Hongda Mao
PY  - 2026
UR  - https://arxiv.org/abs/2608.03112
ID  - 2608.03112
ER  -