Robust pedestrian attribute recognition for an unbalanced dataset using mini-batch training with rarity rate.
Hiroshi FukuiTakayoshi YamashitaYuji YamauchiHironobu FujiyoshiHiroshi MurasePublished in: Intelligent Vehicles Symposium (2016)
Keyphrases
- robust recognition
- recognition accuracy
- partial occlusion
- recognition rate
- training dataset
- automatic recognition
- pattern recognition
- batch mode
- computationally efficient
- recognition algorithm
- varying illumination conditions
- pose invariant face recognition
- geometric hashing
- recognition process
- training set
- object recognition
- feature extraction
- training process
- benchmark datasets
- object detection
- illumination invariance
- character recognition
- supervised learning
- noisy environments
- neural network