Dual-Modal Attention-Enhanced Text-Video Retrieval with Triplet Partial Margin Contrastive Learning

Benchmark Model Rank Results
video-retrieval-on-msr-vtt-1kaDMAE (ViT-B/16)#2text-to-video R@1: 55.5text-to-video R@5: 79.4…
video-retrieval-on-msvdDMAE (ViT-B/32)#9text-to-video R@1: 48.7text-to-video R@5: 78.4…
video-retrieval-on-activitynetDMAE (ViT-B/32)#13text-to-video R@1: 53.4text-to-video R@5: 80.7…
video-retrieval-on-didemoDMAE (ViT-B/32)#16text-to-video R@1: 52.7text-to-video R@5: 79.3…