Online Action Detection On Thumos 14

Metrics

mAP

Results

Performance results of various models on this benchmark

		Paper Title
GateHUB(TimeSformer)	72.5	GateHUB: Gated History Unit with Background Suppression for Online Action Detection
MiniROAD	71.8	MiniROAD: Minimal RNN Framework for Online Action Detection
GateHUB	70.7	GateHUB: Gated History Unit with Background Suppression for Online Action Detection
LSTR	69.5	Long Short-Term Transformer for Online Action Detection
WOAD	67.1	WOAD: Weakly Supervised Online Action Detection in Untrimmed Videos
Colar	66.9	Colar: Effective and Efficient Online Action Detection by Consulting Exemplars
GateHUB (RGB only)	66.5	GateHUB: Gated History Unit with Background Suppression for Online Action Detection
OadTR	65.2	OadTR: Online Action Detection with Transformers
OadTR-b2	64.5	Continual Transformers: Redundancy-Free Attention for Online Inference
CoOadTR-b2	64.4	Continual Transformers: Redundancy-Free Attention for Online Inference
OadTR	64.2	Continual Transformers: Redundancy-Free Attention for Online Inference
OadTR-b1	63.9	Continual Transformers: Redundancy-Free Attention for Online Inference
TRN	62.1	Temporal Recurrent Networks for Online Action Detection
Colar(RGB only)	58.6	Colar: Effective and Efficient Online Action Detection by Consulting Exemplars
CoOadTR-b1	-	Continual Transformers: Redundancy-Free Attention for Online Inference

0 of 15 row(s) selected.