Action Recognition In Videos 2
Action recognition in videos is a task in the field of computer vision and pattern recognition, aiming to identify and classify human actions in video sequences. This task achieves intelligent understanding and parsing of video content by analyzing the spatiotemporal dynamic features of actions and mapping them to a predefined set of action categories, such as running, jumping, or swimming. It has significant application value in areas like surveillance, sports analysis, and human-computer interaction.
ActivityNet
LSTM + Pretrained on YT-8M
AVA v2.1
AVA v2.2
YOWO+LFB*
FS-Something-Something V2-Full
FS-Something-Something V2-Small
ITANet
HMDB-51
Jester (Gesture Recognition)
CPNet Res34, 5 CP
Kinetics-400
Florence
Kinetics-600
Florence
miniSports
G-Blend
NTU RGB+D
2D-3D-Softargmax (RGB only)
PKU-MMD
MMNet
Something-Something V1
Something-Something V2
Sports-1M
THUMOS’14
Single-stream R-C3D (two-way buffer)
UCF101