关键词: 隐马尔可夫模型, 人脸识别, 后处理, 混淆矩阵, 先验模型, 置信度

Abstract: In this paper, the rarely concerned problem of data source in face recognition was investigated, and a novel post-processing HMM-based solution was proposed. Data source problem was firstly empirically investigated through systematically evaluating the eigenfaces sensitivity to variations of pose and illumination by Lambertian reflection model and 3D face model, which revealed that the changes of pose and illumination abruptly degrade the eigenfaces system. This problem is explicitly defined as "data source disaster" for highlighting its significance. Aiming at solving this problem, combining the recognition rate with the analysis of the data sources, two methods were proposed to evaluate the overall performance of specific face recognition approach with its robustness against the low-quality data sources considered. Finally, a post-processing method was proposed to improve the robustness of the recognizer under unconstrained environment. The experimental results have impressively indicated the effectiveness of the proposed post-processing solution to tackle the "data source disaster" problem.

Key words: Hidden Markov Model (HMM), face recognition, post-processing, confusion matrix, priori model, confidence