MixSTE: Seq2seq Mixed Spatio-Temporal Encoder for 3D Human Pose Estimation in Video.
Jinlu ZhangZhigang TuJianyu YangYujin ChenJunsong YuanPublished in: CoRR (2022)
Keyphrases
- spatio temporal
- spatial and temporal
- temporal correlation
- space time
- video representation
- spatial temporal
- video sequences
- human actions
- spatio temporally
- mpeg standard
- multimedia
- video encoding
- temporal domain
- video data
- video encoder
- video content
- real time video
- body parts
- video analysis
- video frames
- video streams
- real time
- temporal segmentation
- action detection
- moving objects
- video retrieval
- video transcoding
- channel bandwidth
- spatial and temporal relationships
- key frames
- rate distortion
- motion estimation
- video codec
- packet loss
- temporal structure
- video compression
- video clips
- low complexity
- temporal information
- video coding
- human body
- bit rate
- image sequences
- computer vision