Action Detection on Charades

Metric: mAP (higher is better)

LeaderboardDataset

Loading chart...

Results

Hide extra data

Sort:

#	Model↕	mAP▼	Extra Data	Paper	Date↕	Code
1	TTM	28.79	No	Token Turing Machines	2022-11-16	Code
2	CTRN	27.8	No	CTRN: Class-Temporal Relational Network for Acti...	2021-10-26	-
3	Coarse-Fine Networks (w/ self-supervised detection pretraining)	26.95	No	Weakly-guided Self-supervised Pretraining for Te...	2021-11-26	Code
4	UniMD+Sync. (RGB+Flow)	26.53	No	UniMD: Towards Unifying Moment Retrieval and Tem...	2024-04-07	Code
5	PDAN (RGB+Flow)	26.5	No	-	-	Code
6	PAT	26.5	No	PAT: Position-Aware Transformer for Dense Multi-...	2023-08-09	-
7	MS-TCT (RGB only)	25.4	No	MS-TCT: Multi-Scale Temporal ConvTransformer for...	2021-12-07	Code
8	3D ResNet-50 + super-events pretrained on AViD	25.2	Yes	AViD Dataset: Anonymized Videos from Diverse Cou...	2020-07-10	Code
9	Coarse-Fine Networks	25.1	No	Coarse-Fine Networks for Temporal Activity Detec...	2021-03-01	Code
10	I3D + biGRU + VS-ST-MPNN	23.7	Yes	Representation Learning on Visual-Symbolic Graph...	2019-05-17	-
11	MLAD (RGB + Flow)	23.7	No	Modeling Multi-Label Action Dependencies for Tem...	2021-03-04	Code
12	3D ResNet-50 pretrained on AViD	23.2	No	AViD Dataset: Anonymized Videos from Diverse Cou...	2020-07-10	Code
13	TGM (RGB+Flow)	22.3	Yes	Temporal Gaussian Mixture Layer for Videos	2018-03-16	Code
14	Super-events (RGB+Flow)	19.41	Yes	Learning Latent Super-Events to Detect Multiple ...	2017-12-05	Code
15	R-C3D	12.4	No	R-C3D: Region Convolutional 3D Network for Tempo...	2017-03-22	Code
16	Sigurdsson et al.	9.6	No	Asynchronous Temporal Fields for Action Recognit...	2016-12-19	Code

#1TTMSOTA
28.79
mAP· 2022-11-16
Token Turing Machines Code
#2CTRNSOTA
27.8
mAP· 2021-10-26
CTRN: Class-Temporal Relational Network for Action Detection
#3Coarse-Fine Networks (w/ self-supervised detection pretraining)
26.95
mAP· 2021-11-26
Weakly-guided Self-supervised Pretraining for Temporal Activity Detection Code
#4UniMD+Sync. (RGB+Flow)
26.53
mAP· 2024-04-07
UniMD: Towards Unifying Moment Retrieval and Temporal Action Detection Code
#5PDAN (RGB+Flow)
26.5
mAP
No paperCode
#6PAT
26.5
mAP· 2023-08-09
PAT: Position-Aware Transformer for Dense Multi-Label Action Detection
#7MS-TCT (RGB only)
25.4
mAP· 2021-12-07
MS-TCT: Multi-Scale Temporal ConvTransformer for Action Detection Code
#83D ResNet-50 + super-events pretrained on AViDSOTA
25.2
mAP· Extra Data· 2020-07-10
AViD Dataset: Anonymized Videos from Diverse Countries Code
#9Coarse-Fine Networks
25.1
mAP· 2021-03-01
Coarse-Fine Networks for Temporal Activity Detection in Videos Code
#10I3D + biGRU + VS-ST-MPNNSOTA
23.7
mAP· Extra Data· 2019-05-17
Representation Learning on Visual-Symbolic Graphs for Video Understanding
#11MLAD (RGB + Flow)
23.7
mAP· 2021-03-04
Modeling Multi-Label Action Dependencies for Temporal Action Localization Code
#123D ResNet-50 pretrained on AViD
23.2
mAP· 2020-07-10
AViD Dataset: Anonymized Videos from Diverse Countries Code
#13TGM (RGB+Flow)SOTA
22.3
mAP· Extra Data· 2018-03-16
Temporal Gaussian Mixture Layer for Videos Code
#14Super-events (RGB+Flow)SOTA
19.41
mAP· Extra Data· 2017-12-05
Learning Latent Super-Events to Detect Multiple Activities in Videos Code
#15R-C3DSOTA
12.4
mAP· 2017-03-22
R-C3D: Region Convolutional 3D Network for Temporal Activity Detection Code
#16Sigurdsson et al.SOTA
9.6
mAP· 2016-12-19
Asynchronous Temporal Fields for Action Recognition Code