FWT-DCT feature based face authentication

This paper presents a new method for face authentication based on Fast Wavelet Transform (FWT) and Discrete Cosine Transform (DCT). When a probe face is not in the database, the near-zero false alarm rate is still a challenge for face authentication. To address this issue, a new near-zero false alarm rate method for face authentication is presented in this paper. The extracted feature in the subspace is called "FWT-DCT feature", and it does not necessarily correspond to features such as eyes, ears and noses. The experiment results demonstrate that the FWT-DCT method achieves near-zero false alarm rate to satisfy the authentication occasion.