MAFA-net: pedestrian detection network based on multi-scale attention feature aggregation.
Hao XiaHonglin WanJiayu OuJun MaXinyao LvChengjie BaiPublished in: Appl. Intell. (2022)
Keyphrases
- pedestrian detection
- multiscale
- histograms of oriented gradients
- human body
- object detection
- cascade classifier
- human detection
- benchmark datasets
- pedestrian tracking
- detection rate
- image features
- occlusion handling
- natural images
- face detection
- visual attention
- crowded scenes
- feature selection
- computer vision
- computational cost
- action recognition
- feature vectors
- viewpoint
- computational complexity
- adaboost classifier
- three dimensional