SOANets: Encoder-decoder based Skeleton Orientation Alignment Network for White Cane User Recognition from 2D Human Skeleton Sequence.
Naoki NishidaYasutomo KawanishiDaisuke DeguchiIchiro IdeHiroshi MuraseJun PiaoPublished in: VISIGRAPP (5: VISAPP) (2020)
Keyphrases
- skeleton extraction
- shape representation
- binary images
- user interface
- human users
- object recognition
- low complexity
- recognition rate
- wyner ziv video coding
- human activities
- individual user
- video codec
- medial axis
- recommender systems
- action recognition
- error control
- distributed video coding
- shape analysis
- gait recognition
- wireless sensor networks
- character recognition
- wyner ziv
- motion compensation
- decoding process
- video coding scheme
- video coding
- rate distortion