Host: The Institute of Image Electronics Engineers of Japan
Co-host: Osaka Institute of Technology, Information Processing Society of Japan
We propose a new face detection method based on local self correlation using wavelet transform. Our method applies wavelet transform for face images and divides the face images into a number of domains. Moreover we calculate the most similar domain for each domain and calculate statistical probability using a number of face images. Then, we detect the face region using this statistical probability. Experimental results show that our proposed method achieve more high detection rate than other methods.