| 非限制环境下的低秩协同人脸性别识别研究 |
| |
| 引用本文： | 孙宁,郭行,刘佶鑫,韩光.非限制环境下的低秩协同人脸性别识别研究[J].电子测量与仪器学报,2016,30(11):1790-1800. |
| |
| 作者姓名： | 孙宁 郭行 刘佶鑫 韩光 |
| |
| 作者单位： | 南京邮电大学宽带无线通信技术教育部工程研究中心 南京210003 |
| |
| 基金项目： | 国家自然科学基金(61471206;61302156)，江苏省自然科学基金(BK20141428 |
| |
| 摘 要： | 人脸性别识别是计算机视觉和机器学习的热门研究课题,但目前大多数的人脸性别识别算法对自然环境下的图像进行识别的效果并不理想,识别正确率与实际应用差距较大。采用低秩分解和协同表示来提高人脸性别识别的正确率和鲁棒性。在提取特征前采用低秩分解配准未对齐的图像并降低污染和噪声的影响,使得提取的特征能够很好地反映图像原有的信息。识别环节采用稀疏表示的改进算法—协同表示,其使用l2范数替代l1范数优化问题易于求解。在实验中,选用AR,CAS-PEAL和You Tube三种数据库对算法进行测试,结果表明本算法性能与其他主流算法相比有明显优势。
|
| 关 键 词： | 人脸性别识别 低秩分解 协同表示 |
Research on low rank and collaborative representation for facial gender recognition in unconstrained environment |
| |
| Sun Ning,Guo Hang,Liu Jixin and Han Guang.Research on low rank and collaborative representation for facial gender recognition in unconstrained environment[J].Journal of Electronic Measurement and Instrument,2016,30(11):1790-1800. |
| |
| Authors: | Sun Ning Guo Hang Liu Jixin and Han Guang |
| |
| Affiliation: | Engineering Research Center of Wideband Wireless Communication Technology, Ministry of Education, Nanjing University of Posts and Telecommunications, Nanjing 210003, China,Engineering Research Center of Wideband Wireless Communication Technology, Ministry of Education, Nanjing University of Posts and Telecommunications, Nanjing 210003, China,Engineering Research Center of Wideband Wireless Communication Technology, Ministry of Education, Nanjing University of Posts and Telecommunications, Nanjing 210003, China and Engineering Research Center of Wideband Wireless Communication Technology, Ministry of Education, Nanjing University of Posts and Telecommunications, Nanjing 210003, China |
| |
| Abstract: | Facial gender recognition is an attractive topic in the field of computer vision and machine learning. But, most of the existing facial gender recognition methods are always suffering from the problem of weak robustness when working on the unconstrained environment. For this reason, the low rank decomposition and collaborative representation are exploited to raise the precision and the robustness of the facial gender recognition algorithm. The low rank decomposition to minimize the negative effect caused by image corruption and various face poses is used in this paper. In the phase of recognition, the collaborative representation mechanism, which substitutes l1 norm regularization with l2 norm regularization to enhance the discriminant power, decreases the computational cost of the gender recognition system and solves the small sample size problem of facial gender recognition. The experiments evaluate the proposed method on the several benchmarks. The result demonstrates that the proposed method is able to achieve better performance than that of the current state of art facial gender recognition approaches. |
| |
| Keywords: | facial gender recognition low-rank decomposition collaborative representation |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
| 点击此处可从《电子测量与仪器学报》浏览原始摘要信息 |
|
点击此处可从《电子测量与仪器学报》下载全文 |
|
