IVTP: Instruction-Guided Visual Token Pruning for Large Vision-Language Models.

Kai Huang, Hao Zou, Ye Xi, Bochen Wang, Zhen Xie, Liang Yu

Browse the full ECCV paper archive.