List your company

Video Question Answering on EgoSchema 2023 (test)

58.55AccuracyGSMT-CLIP
33.912440.308746.70553.1013Jan 31, 2026

Evaluation results

MethodLinks
GSMT-CLIPPre-trained model=CLIP (ViT/B-32)2026.0158.55
GSMT-AIOPre-trained model=All-in-One-Base2026.0158.28
MIST-CLIPPre-trained model=CLIP (ViT/B-32)2026.0156.42
MIST-AIOPre-trained model=All-in-One-Base2026.0156.27
VideoReCap2026.0150.23
EgoVLP2026.0134.86