List your company

Action Recognition on Something-Else

74.5Top-1 AccuracyObjectViViT
35.70845.77955.8565.921Jun 20, 2023Nov 19, 2023Apr 20, 2024Sep 20, 2024Feb 20, 2025Jul 23, 2025Dec 23, 2025

Evaluation results

MethodLinks
ObjectViViTn. FLOPs=1.06x, Temporal views=2, Spatial crops=3, Token sampling percentage=50%2023.0674.5—
MFormer w. ObjectLearnern. FLOPs=1.04x, Temporal views=1, Spatial crops=32023.0673.6—
ObjectViViTn. FLOPs=1.06x, Temporal views=1, Spatial crops=32023.0672—
VideoMAE w. ORVIT (reproduced)n. FLOPs=1.06x, Temporal views=1, Spatial crops=32023.0671.6—
MFormer w. ORVITn. FLOPs=1.09x, Temporal views=1, Spatial crops=32023.0669.7—
CACNFModality=RGB-based and multimodal fusion-based, Evaluation Protocol=Oracle2025.1267.190.4
VideoMAE (our baseline)n. FLOPs=1x (181), Temporal views=1, Spatial crops=32023.0667—
MFormern. FLOPs=1x (370), Temporal views=1, Spatial crops=32023.0662.8—
SFIModality=RGB-based and multimodal fusion-based, Evaluation Protocol=Oracle2025.1259.685.8
StudentBackbone=Swin-T, Input Modality=RGB, lambda=1, gamma=302025.1259.186.1
STLTModality=Object detections-based, Evaluation Protocol=Oracle2025.125986
OmnivoreBackbone=Swin-T, Input Modality=RGB2025.1258.384.9
CACNFModality=RGB-based and multimodal fusion-based, Evaluation Protocol=Obj. predictions2025.1256.982.5
STIN + I3DModality=RGB-based and multimodal fusion-based, Evaluation Protocol=Oracle2025.1254.679.4
STRGModality=RGB-based and multimodal fusion-based, Evaluation Protocol=Obj. predictions2025.1252.378.3
STINModality=Object detections-based, Evaluation Protocol=Oracle2025.1251.479.3
STIN + I3DModality=RGB-based and multimodal fusion-based, Evaluation Protocol=Obj. predictions2025.1248.272.6
I3DModality=RGB-based and multimodal fusion-based, Evaluation Protocol=Obj. predictions2025.1246.872.2
I3DModality=RGB-based and multimodal fusion-based, Evaluation Protocol=Oracle2025.1246.872.2
SFIModality=Object detections-based, Evaluation Protocol=Oracle2025.1244.174
STLTModality=Object detections-based, Evaluation Protocol=Obj. predictions2025.1241.667.9
STINModality=Object detections-based, Evaluation Protocol=Obj. predictions2025.1237.262.4