List your company

Video Action Classification on COIN

95.3Top-1 AccPEcoreG
6.79629.77352.7575.727Mar 31, 2023Aug 11, 2023Dec 23, 2023May 5, 2024Sep 16, 2024Jan 28, 2025Jun 11, 2025

Evaluation results

MethodLinks
PEcoreGParameters=1.9B, Evaluation Protocol=Attentive probe2025.0695.3
SigLIP2Parameters=1.2B, Evaluation Protocol=Attentive probe2025.0695.1
InternVideo2_s2-1BParameters=1B, Evaluation Protocol=Attentive probe2025.0693.8
HERMES2024.0893.5
MA-LMM2024.0893.2
MA-LMM2024.0493.2
V-JEPA 2 ViT-g384Parameters=1B, Input format=32x8x3 frames, Input Resolution=384x3842025.0691.1
S52024.0890.8
S52024.0490.8
DINOv2Parameters=1.1B, Evaluation Protocol=Attentive probe2025.0690.7
V-JEPA 2 ViT-gParameters=1B, Input format=32x8x3 frames2025.0690.7
D-Sprv.2024.0490
TranS4mer2024.0889.2
VIS4mer2024.0888.4
ViS4mer2024.0488.4
V-JEPA 2 ViT-HParameters=600M, Input format=32x8x3 frames2025.0687.9
V-JEPA ViT-HParameters=600M, Evaluation Protocol=Attentive probe2025.0687.1
V-JEPA 2 ViT-LParameters=300M, Input format=32x8x3 frames2025.0686.8
TSN2024.0473.4
ProcedureVRLPretraining Supervision=Unsupervised: ASR, Pretraining Dataset=HT100M, Protocol=Fine-tuning2023.0356.9
DistantSupPretraining Supervision=Unsupervised: ASR + wikiHow, Pretraining Dataset=HT100M, Protocol=Fine-tuning2023.0354.1
TimeSformerPretraining Supervision=Supervised: action labels, Pretraining Dataset=Kinetics, Protocol=Fine-tuning2023.0348.3
DistantSup†Pretraining Supervision=Unsupervised: ASR + wikiHow, Pretraining Dataset=HT100M, Protocol=Fine-tuning2023.0346.6
TimeSformerPretraining Supervision=Unsupervised: ASR w. MIL-NCE [39], Pretraining Dataset=HT100M, Protocol=Fine-tuning2023.0346.5
CLIPPretraining Supervision=Unsupervised: captions, Pretraining Dataset=CLIP400M, Protocol=Fine-tuning2023.0345.9
VideoCLIPPretraining Supervision=Unsupervised: ASR, Pretraining Dataset=HT100M, Protocol=Fine-tuning2023.0339.4
S3DPretraining Supervision=Unsupervised: ASR w. MIL-NCE [39], Pretraining Dataset=HT100M, Protocol=Fine-tuning2023.0337.5
TSN (RGB+Flow)Pretraining Supervision=Supervised: action labels, Pretraining Dataset=Kinetics, Protocol=Fine-tuning2023.0336.5
SlowFastPretraining Supervision=Supervised: action labels, Pretraining Dataset=Kinetics, Protocol=Fine-tuning2023.0332.9
ClipBERTPretraining Supervision=Supervised: captions, Pretraining Dataset=COCO+VG, Protocol=Fine-tuning2023.0330.8
ProcedureVRLPretraining Supervision=Unsupervised: ASR, Pretraining Dataset=HT100M, Protocol=Zero-shot2023.0316.6
CLIPPretraining Supervision=Unsupervised: captions, Pretraining Dataset=CLIP400M, Protocol=Zero-shot2023.0314.8
DistantSup†Pretraining Supervision=Unsupervised: ASR + wikiHow, Pretraining Dataset=HT100M, Protocol=Zero-shot2023.0310.2