Cost-effective real-time recognition for human emotion-age-gender using deep learning with normalized facial cropping preprocess.
Ta-Te LuSheng-Cheng YehChia-Hui WangMin-Rou WeiPublished in: Multim. Tools Appl. (2021)
Keyphrases
- cost effective
- deep learning
- real time
- low cost
- facial expressions
- human emotion
- unsupervised learning
- facial expression recognition
- object recognition
- cost effectiveness
- machine learning
- pattern recognition
- unsupervised feature learning
- mental models
- weakly supervised
- facial images
- human faces
- image content
- feature extraction
- face images
- d objects
- image features
- probabilistic model
- active learning
- viewpoint
- multiscale
- computer vision