为了提高跨视角步态识别的准确率,充分提取步态中的时间信息,提出了一种基于步态序列的跨视角步态识别模型,该模型利用编码器,并引入三元组损失函数,以此提取步态序列的特征,通过使用生成器与判别器,以及连续帧判别损失对编码器进行修正,确保提取具有时间信息的有效步态特征.针对CASIA-B数据集和OU-MVLP数据集,对提出的方法进行了实验研究,且与卷积神经网络和步态能量图方法进行了实验比较,验证了提出方法的有效性.
In order to improve the accuracy of cross-view gait recognition and fully extract the temporal information in gait, we proposed a cross-view gait recognition model based on gait image sequence. The model combines encoder and triplet-loss to extract the features from gait sequences. Meanwhile, using a generator and discriminators, molel optimizes the encoder according the discriminant loss so that extracts the temporal information in gait. In order to validate the effectiveness of the proposed method, we conducted experimental studies on CASIA-B and OU-MVLP datasets. Moreover, we also compare with other state-of-arts methods which are based on convolutional neural network and gait energy image.
[1] ARIYANTO G, NIXON M S. Model-based 3D gait biometrics[Z]. International Joint Conference on Biometrics, Washington, DC, USA, 2011. DOI:10.1109/ijcb.2011.6117582.
[2] BODOR R, DRENNER A, FEHR D, et al. View-independent human motion classification using image-based reconstruction[J]. Image and Vision Computing, 2009, 27(8): 1194-1206. DOI:10.1016/j.imavis.2008.11.008.
[3] GOFFREDO M, BOUCHRIKA I, CARTER J N, et al. Self-calibrating view-invariant gait biometrics[J]. IEEE Transactions on Systems, Man, and Cybernetics, Part B(Cybernetics), 2010, 40(4): 997-1008. DOI:10.1109/tsmcb.2009.2031091.
[4] MAKIHARA Y, SUZUKI A, MURAMATSU D, et al. Joint intensity and spatial metric learning for robust gait recognition[Z]. IEEE Conference on Computer Vision and Pattern Recognition, Honolulu, 2017. DOI:10.1109/cvpr.2017.718.
[5] KUSAKUNNIRAN W, WU Q, LI H D, et al. Multiple views gait recognition using View Transformation Model based on optimized Gait Energy Image[Z]. IEEE 12th International Conference on Computer Vision Workshops, Kyoto, Japan, 2009. DOI:10.1109/iccvw.2009.5457587.
[6] HU M D, WANG Y H, ZHANG Z X, et al. View-invariant discriminative projection for multi-view gait-based human identification[J]. IEEE Transactions on Information Forensics and Security, 2013, 8(12): 2034-2045. DOI:10.1109/tifs.2013.2287605.
[7] YU S Q, CHEN H F, REYES E B G, et al. GaitGAN: invariant gait feature extraction using generative adversarial networks[Z]. IEEE Conference on Computer Vision and Pattern Recognition Workshops, Honolulu, 2017. DOI:10.1109/cvprw.2017.80.
[8] YAN C, ZHANG B L, COENEN F. Multi-attributes gait identification by convolutional neural networks[Z]. 8th International Congress on Image and Signal Processing, Shenyang, 2015. DOI:10.1109/cisp.2015.7407957.
[9] WU Z F, HUANG Y Z, WANG L. Learning representative deep features for image set analysis[J]. IEEE Transactions on Multimedia, 2015, 17(11): 1960-1968. DOI:10.1109/tmm.2015.2477681.
[10] CASTRO F M, MARÍN-JIMÉNEZ M J, GUIL N, et al. Automatic learning of gait signatures for people identification[M] //Advances in Computational Intelligence, Cham: Springer International Publishing, 2017: 257-270. DOI:10.1007/978-3-319-59147-6_23
[11] WOLF T, BABAEE M, RIGOLL G. Multi-view gait recognition using 3D convolutional neural networks[Z]. IEEE International Conference on Image Processing, Phoenix, 2016. DOI:10.1109/icip.2016.7533144.
[12] FENG Y, LI Y C, LUO J B. Learning effective Gait features using LSTM[Z]. 23rd International Conference on Pattern Recognition, Cancun. 2016.DOI:10.1109/icpr.2016.7899654.
[13] TONG S B, LING H F, FU Y Z, et al. Cross-view gait identification with embedded learning[Z]. Thematic Workshops of ACM Multimedia 2017 - Thematic Workshops'17, Mountain Uiew, CA,USA. DOI:10.1145/3126686.3126753.
[14] HUANG Y Y, ZHANG J F, ZHAO H H, et al. Attention-based network for cross-view gait recognition[M] //Neural Information Processing, Cham: Springer International Publishing, 2018: 489-498. DOI:10.1007/978-3-030-04239-4_44.
[15] GOODFELLOW I J, POUGET-ABADIE J,MIRZA M,et al.Generative adversarial networks[J].Advances in Neural Information Processing Systems. 2014, 3:2672-2680. DOI:10.1145/3422622.
[16] RADFORD A, METZ L, CHINTALA S. Unsupervised representation learning with deep convolutional generative adversarial networks[Z].4th International Conference on Learning Representation ICLR 2016 San Juan, Puerto Rico.
[17] MIRZA M, OSINDERO S. Conditional generative adversarial nets[J].Computer Science,2014:2672-2680.
[18] LEDIG C, THEIS L, HUSZAR F, et al. Photo-realistic single image super-resolution using a generative adversarial network[Z]. Conference on Computer Vision and Pattern Recognition(CVPR), Honolulu, 2017. DOI:10.1109/cvpr.2017.19.
[19] YEH R A, CHEN, LIM T Y, et al. Semantic image inpainting with deep generative models[Z]. IEEE Conference on Computer Vision and Pattern Recognition(CVPR), Honolulu, 2017. DOI:10.1109/cvpr.2017.728.
[20] SCHROFF F, KALENICHENKO D, PHILBIN J. FaceNet: A unified embedding for face recognition and clustering[Z]. IEEE Conference on Computer Vision and Pattern Recognition(CVPR), Boston, 2015. DOI:10.1109/cvpr.2015.7298682.
[21] HE Y W, ZHANG J P, SHAN H M, et al. Multi-task GANs for view-specific feature learning in gait recognition[J]. IEEE Transactions on Information Forensics and Security, 2019, 14(1): 102-113. DOI:10.1109/tifs.2018.2844819.
[22] YU S Q, CHEN H F, WANG Q, et al. Invariant feature extraction for gait recognition using only one uniform model[J]. Neurocomputing, 2017, 239: 81-93. DOI:10.1016/j.neucom.2017.02.006. (
