Automatic human age estimation system using fusion of local and global features
Subhash Chand Agrawal, Rajesh Kumar Tripathi, Anand Singh Jalal · 2017
Human beings can easily estimate the person's age from a facial image but this ability is not promising in machines. To address this problem, we propose a novel scheme of age estimation using fusion of local and global features. Local feature, Gabor filter is applied after detecting four parts (Forehead, left cheek, right cheek and mouth) around face and eye regions. Global feature, Histogram of Oriented Gradients (HOG) is applied on cropped face. Experimental results show that combination of local and global features outperforms many previous approaches.