Pixel-wise human motion segmentation using learning vector quantization.
Mochamad HariadiAkio HaradaTakafumi AokiTatsuo HiguchiPublished in: ICARCV (2002)
Keyphrases
- pixel wise
- human motion
- learning vector quantization
- pixel level
- human body
- spatial information
- video sequences
- image sequences
- spatio temporal
- appearance model
- background subtraction
- foreground objects
- template matching
- segmentation algorithm
- neural network
- self organizing maps
- matching algorithm
- object detection
- viewpoint
- computer vision