MVP: Multimodality-Guided Visual Pre-training.

Longhui Wei, Lingxi Xie, Wengang Zhou, Houqiang Li, Qi Tian

Browse the full ECCV paper archive.