MAD: A Scalable Dataset for Language Grounding in Videos from Movie Audio Descriptions

Benchmark Model Rank Results
natural-language-moment-retrieval-on-madCLIP#7R@1,IoU=0.1: 6.57R@1,IoU=0.3: 3.13R@1,IoU=0.5: 1.39…
natural-language-moment-retrieval-on-madVLG-Net#9R@1,IoU=0.1: 3.50R@1,IoU=0.3: 2.63R@1,IoU=0.5: 1.61…
natural-language-moment-retrieval-on-madRandom Chance#10R@1,IoU=0.1: 0.09R@1,IoU=0.3: 0.04R@1,IoU=0.5: 0.01…