作者位次 #3/7SIGIR 2026Oral· Information Retrieval (Full)
Seeing the Whole Through the Parts: Discovering Objects through Semantic Part Mining in Weak Supervision
Shucheng Li · Weixuan Xu · Le Jiang · Hao Wu · Fengyuan Xu · Fan Wu · Feng Lyu
摘要
Weakly Supervised Object Detection (WSOD) is fundamentally limited by instance ambiguity, manifesting as either part domination (focusing on discriminative fragments) or merged detection (confusing objects with context). Unlike existing approaches that rely on object-level patterns, we draw inspi...