MViT-B-24, 32x3 (Kinetics-600 pretraining)

Reported on 3 benchmarks across 3 tasks · 1 paper

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Computer Vision1 result

VideoonCharades
MAP· 2021-04-22
47.7
best: 66.3 (TokenLearner)
Multiscale Vision Transformers arXiv:2104.11227

Robots1 result

Activity RecognitiononAVA v2.2
mAP· 2021-04-22
28.7
best: 45.1 (LART (Hiera-H, K700 PT+FT))
Multiscale Vision Transformers arXiv:2104.11227

Time Series1 result

Action RecognitiononAVA v2.2
mAP· 2021-04-22
28.7
best: 45.1 (LART (Hiera-H, K700 PT+FT))
Multiscale Vision Transformers arXiv:2104.11227