Enhancing Face Recognition Using PC-SIFTTechnique: A Hybrid Approach

Deepti Ahlawat, Vijay Nehra
2017 International Journal of Signal Processing, Image Processing and Pattern Recognition  
In this investigation, a hybrid approach using phase congruency (PC) and Shift Invariant Feature Transform (SIFT) for face recognition is presented. The present study exploits the unique advantages of both phase congruency and Shift Invariant Feature Transform and makes it an effective feature extraction technique. Support Vector Machine (SVM) is used for the classification. The effectiveness of the present work is verified and compared by using another classifier i.e. K-Means. The results of
more » ... s. The results of this study shows that PC -SIFT are more robust to expression variations and shows better results than other methods. The present approach achieves good recognition accuracy on Japanese Female Facial Expression (JAFFE) and Cohn Kanade (CK) database. From the experiments it is found that PC-SIFT outperforms SIFT technique and provide an average recognition accuracy of 92.4% and 91.5% respectively for different expression for JAFFE and CK datasets. The proposed technique has been compared with the state-of-art techniques and it is observed that proposed approach gives better results than existing techniques. 11 Keeping in view the literature survey and above facts, the objective of present investigation is to utilize PC-SIFT feature extraction technique for face recognition and compare the results using SVM and K-Means classifier on JAFFE and CK datasets. After a brief overview of the literature survey of the present investigation, in the next section, material and method adopted for carrying out research is presented.
doi:10.14257/ijsip.2017.10.9.02 fatcat:tme7zabjg5hdzpivssg3qi5bdi