1 paper
Yi Zhao, Siqi Wang, Qiqun Geng +2
Vision-Language Model (VLM)-based assistance is reshaping independence for blind and low-vision (BLV) users, yet current tools fail in dynamic settings. While request-response arch…