Guozhi Tang

Reinforcement Learning (1)

visual information extraction (2) text classification (1) weakly supervised learning (1) knowledge distillation (1) image captioning (1) document understanding (1) vision-language alignment (1) sequence tagging (1) end-to-end learning (1) multimodal large language model (1) vision encoder (1) mutual learning (1) reasoning capability (1) text detection (1) optical character recognition (1) text recognition (1) text spotting (1) verifiable reward (1) multimodal feature fusion (1) expert prompt (1)