Hello,
Does anybody know which classifier has high performance in Low resolution frontal face recognition under SURF + HOG feature set.
Suggestions are welcomed.
Thank you.
Would you recognize the people of that there is a human face in the image?
@ Cedric Lemaitre... Thanks for your response, I appreciate. It's the Recognition of human face in the image. I am working with Labelled Faces in the Wild LFW data sets...
SVM usually gives better performance in this kind of feature set.. However, I would recommend starting with K-means then move to KNN for understanding the essence of classification algorithm. All in all, depending on the image set, in general, SVM gives better performance than these algorithms provided you have some labeled training data. Good luck.
@ Hameed R. Farhan. Thanks so much, I appreciate your suggestions.
@ Cedric Lemaitre. Yes, I am using Viola Jones algorithm to detect faces on LFW data sets. What I am actually concerned about it the best Classification technique to match Extracted features with high accuracy. I intend using a Multi-modal Feature Extraction technique (i.e a combination of 2 or more feature extraction techniques) to obtain a more discriminative information of my Training set. Should you have any suggestion, kindly share with me. I am new to Face recognition and currently working on "Low resolution Face recognition with Occlusion handling". Thank you
Why not use facial landmark method?
It is fast and reliable.
You could try : http://dlib.net/dnn_face_recognition_ex.cpp.html
- Badges
- Science topic
- Similar topics
- Analytical Chemistry
- Extraction