Vehicle counting in crowded scenes with multi-channel and multi-task convolutional neural networks.
Maojin SunYan WangTeng LiJing LvJun WuPublished in: J. Vis. Commun. Image Represent. (2017)
Keyphrases
- multi channel
- multi task
- crowded scenes
- pedestrian detection
- convolutional neural networks
- human body
- human detection
- learning tasks
- object detection
- spatial and temporal
- multi class
- detection rate
- ground plane
- learning problems
- benchmark datasets
- space time
- behavior analysis
- transfer learning
- feature selection
- object tracking
- camera calibration
- image classification
- complex scenes
- video sequences