一种鲁棒稀疏表示的单样本人脸识别算法

稀疏表示在人脸识别问题上取得了非常优秀的识别结果，但在单样本条件下，算法性能下降严重。为提高单样本条件下稀疏表示的应用能力，提出一种鲁棒稀疏表示单样本人脸识别算法(RSR)。通过使用每张人脸图像创建一组位置图像，扩充每个对象训练样本，并利用L2,1范数约束，保证RSR选择正确对象的位置图像。在AR和Extended Yale B人脸数据库上进行评测，实验结果表明RSR能够有效处理存在遮挡或光照变化的人脸图像，获得了较好的单样本人脸识别准确率，具有很强的鲁棒性。

Single Sample Face Recognition based on Robust Sparse Representation

Sparse Representation(SR) has successfully addressed the face recognition problem with sufficient training images of each gallery subject, however, its performance will deteriorate much for single sample face recognition(SSFR). To improve the generalization ability of SR for SSFR, a Robust Sparse Representation(RSR) method is proposed. By creating a set of position images for each training picture to expand the training samples of each gallery subject and using L2,1 norm to prompt RSR to select the correct position images. Finally, the proposed method is evaluated on AR and Extended Yale B face databases, the experimental results demonstrate the effectiveness and robustness of the approach.