Jiajun Cao

token pruning (2) knowledge distillation (1) autonomous driving (1) efficient computing (1) visual attention (1) vision language model (1) cross-modal alignment (1) low-rank adaptation (1) mixture of expert (1) vision-language model (1) latency reduction (1) visual encoder (1) vision language action (1) modality bia (1) gradient-based filtering (1) plug and play (1) foreground reconstruction (1) intra-modal learning (1) multimodal keyphrase generation (1) fine-grained feature extraction (1)