Improved softmax loss for deep learning-based face and expression recognition.
Jiancan ZhouXi JiaLinlin ShenZhenkun WenZhong MingPublished in: Cogn. Comput. Syst. (2019)
Keyphrases
- expression recognition
- deep learning
- facial expressions
- facial expression recognition
- unsupervised learning
- face identification
- face images
- face detection
- machine learning
- facial features
- human faces
- mental models
- object detection
- video sequences
- facial images
- weakly supervised
- data points
- maximum likelihood
- data sets
- viewpoint
- feature extraction
- pose variations
- image segmentation
- computer vision