VLPD: Context-Aware Pedestrian Detection via Vision-Language Semantic Self-Supervision.

Mengyin Liu, Jie Jiang, Chao Zhu, Xu-Cheng Yin

Browse the full CVPR paper archive.