【摘要】 人脸识别一直是模式识别和机器视觉领域研究热点。受姿态变化和分辨率低影响,传统方法对伴随多姿态低分辨率人脸图像识别精度较低。为充分挖掘姿态变化带来的非线性问题,提出将深度信念网络与极限学习机相结合来识别低分辨率多姿态人脸图像。该方法将低分辨率和对应高分辨率图像作为深层网络结构输入数据,学习高低分辨率图像间流行假设的点对联系以提取特征进行分类识别。实验结果表明,所提方法相比于其他方法具有识别率高、分类时间短等优点。更多还原

【Abstract】 Face recognition has been a hot topic in the field of pattern recognition and machine vision. Affected by pose variations and low resolution, the recognition accuracy of the traditional method is very low. In order to exploit the non-linear contact of pose variations, a deep network structure that combines deep belief networks and extreme learning machines is proposed. We extract the features by learning the point-pair relationship of the manifold assumption from HR manifold and LR manifold by sending both HR images and LR images to a deep architecture. Finally, the experimental results show that the proposed method has advantages of high recognition rate and short classification time compared with other methods.更多还原