Skip to content

Latest commit

 

History

History
17 lines (16 loc) · 1.04 KB

File metadata and controls

17 lines (16 loc) · 1.04 KB

Dissimilarity-based_nearest_neighbor_classifier-for-single-sample_face-_ecognition

Dissimilarity-based nearest neighbor classifier for single-sample face recognition Zhengqi Zhang, Li Zhang, Meng Zhang A dissimilarity-based nearest neighbor classifier for face recognition single sample Yasman Safari Pattern recognition lesson Professor: Dr. Ebrahimpour Abstract: In face recognition tasks with a single training sample, nearest neighbor class is one of the famous methods The idea is the simplicity of implementation. Choosing the right distance measurement criteria in these issues due to the effects of light and Other photography issues should be done with more accuracy because traditional methods such as Euclidean distance are of good accuracy do not have In this paper, a measure of dissimilarity or distance was introduced. Training and test samples in the form of patches The introduced distance criterion was applied to these patches from the segments of each photo. Finally, the accuracy is class The nearest neighbor was calculated. Introduction