al Neural Network, I/O-ACNN)分析不同的网络结构和损失函数对步态识别准确率的影响, 使用步态能量图作为模型输入.对于损失函数, 对比度量学习中常用的二元组损失(Contrastive Loss)[46]和三元组损失(Triplet Loss)[47].二元组损失定义为[11]
其中:δ i, j为指示函数, 表示训练集中第i个和第j个样本是否具有相同的身份, 如果相同值为1, 否则为0; di, j为特征之间的欧氏距离.该损失函数对于相同身份的样本最小化特征之间的距离, 对于不同身份的样本, 令特征之间的距离大于某一个阈值margin.
三元组损失定义为[11]
其中, da, n表示具有不同身份的特征之间的距离, da, p表示具有相同身份的特征之间的欧氏距离.为了分析步态识别中类内和类间样本空间不对齐问题对步态识别性能的影响, 文献[11]方法对比2种模型结构:1)在模型的输入层上进行求绝对差运算, 即在输入层进行特征融合; 2)直接在网络结构的高层学习特征表示.文献[11]方法的优势在于提出融合不同网络结构的方法, 可以同时适用于验证和辨别任务.
HAN J, BHANU B. Individual Recognition Using Gait Energy Image. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2005, 28(2): 316-322. [本文引用:4]
钟兴志, 王晨升, 刘丰, 等. 步态识别综述. 软件, 2013, 34(4): 160-164.
(ZHONG X Z, WANG C S, LIU F, et al. A Survey of Gait Recognition. Software, 2013, 34(4): 160-164. ) [本文引用:2]
王科俊, 侯本博. 步态识别综述. 中国图象图形学报, 2007, 12(7): 1152-1160.
(WANG K J, HOU B B. A Survey of Gait Recognition. Journal of Image and Graphics, 2007, 12(7): 1152-1160. ) [本文引用:2]
LÜ Z W, XING X L, WANG K J, et al. Class Energy Image Analysis for Video Sensor-Based Gait Recognition: A Review. Sensors, 2015, 15(1): 932-964. [本文引用:2]
贲晛烨, 徐森, 王科俊. 行人步态的特征表达及识别综述. 模式识别与人工智能, 2012, 25(1): 71-81.
(BEN X Y, XU S, WANG K J, et al. Review on Pedestrian Gait Feature Expression and Recognition. Pattern Recognition and Artificial Intelligence, 2012, 25(1): 71-81. ) [本文引用:2]
陈昌由, 张军平. 步态识别的特征提取综述. 计算机研究与发展, 2007, 44(S2): 361-365. (CHEN C Y, ZHANG J P. A Survey of Feature Extraction on Gait Recognition. Journal of Computer Research and Development, 2007, 44(S2): 361-365 ) [本文引用:2]
HE K M, GEORGIA G, PIOTR D, et al. Mask R-CNN // Proc of the IEEE International Conference on Computer Vision. Washington, USA: IEEE, 2017: 2980-2988. [本文引用:1]
BOYKOV Y Y, JOLLY M P. Interactive Graph Cuts for Optimal Boundary and Region Segmentation of Objects in N-D Images // Proc of the 8th IEEE International Conference on Computer Vision. Washington, USAUSA: IEEE, 2001: 105-112. [本文引用:1]
BULAT A, TZIMIROPOULOS G. Human Pose Estimation via Convo-lutional Part Heatmap Regression // Proc of the European Confe-rence on Computer Vision. Berlin, Germany: Springer, 2016: 717-732. [本文引用:2]
CAO Z, SIMON T, WEI S E, et al. Realtime Multi-person 2D Pose Estimation Using Part Affinity Fields // Proc of the IEEE Conference on Computer Vision and Pattern Recognition. Washington, USA: IEEE, 2017: 1302-1310. [本文引用:2]
TAKEMURA N, MAKIHARA Y, MURAMATSU D, et al. On Input/Output Architectures for Convolutional Neural Network-Based Cross-View Gait Recognition. IEEE Transactions on Circuits and Systems for Video Technology, 2018. DOI DOI:10.1109/TCSVT.2017.2760835. [本文引用:8]
IWAMA H, OKUMURA M, MAHIKARA Y, et al. The OU-ISIR Gait Database Comprising the Large Population Dataset and Performance Evaluation of Gait Recognition. IEEE Transactions on Information Forensics and Security, 2012, 7(5): 1511-1521. [本文引用:1]
MAKIHARA Y, SUZUKI A, MURAMATSU D, et al. Joint Intensity and Spatial Metric Learning for Robust Gait Recognition // Proc of the IEEE Conference on Computer Vision and Pattern Re-cognition. Washington, USA: IEEE, 2017: 6786-6796. [本文引用:1]
YU S Q, TAN D L, TAN T N. A Framework for Evaluating the Effect of View Angle, Clothing and Carrying Condition on Gait Recognition // Proc of the IEEE International Conference on Pa-ttern Recognition. Washington. USA: IEEE, 2006: 441-444. [本文引用:2]
SARKAR S, PHILLIPS P J, LIU Z Y, et al. The HumanID Gait Challenge Problem: Data Sets, Performance, and Analysis. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2005, 27(2): 162-177. [本文引用:1]
NIXON M S, CATER J N, CUNADO D, et al. Automatic Gait Recognition. Motion Analysis and Tracking, 1999, 7(2): 3-6. [本文引用:2]
WANG L, TAN T N, NING H Z, et al. Silhouette Analysis-Based Gait Recognition for Human Identification. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2003, 25(12): 1505-1518. [本文引用:1]
WANG L, TAN T N, HU W M, et al. Automatic Gait Recognition Based on Statistical Shape Analysis. IEEE Transactions on Image Processing, 2003, 12(9): 1120-1131. [本文引用:1]
WANG L, NING H Z, TAN T N, et al. Fusion of Static and Dynamic Body Biometrics for Gait Recognition. IEEE Transactions on Circuits and Systems for Video Technology, 2004, 14(2): 149-158. [本文引用:1]
HAYFRON-ACQUAH J B, NIXON M S, CARTER J N. Automa-tic Gait Recognition by Symmetry Analysis. Pattern Recognition Letters, 2003, 24(13): 2175-2183. [本文引用:1]
BOBICK A F, JOHNSON A K. Gait Recognition Using Static, Activity-Specific Parameters // Proc of the IEEE Conference on Computer Vision and Pattern Recognition. Washington, USA: IEEE, 2001: 423-430. [本文引用:1]
BASHIR K, XIANG T, GONG S. Gait Recognition Using Gait Entropy Image // Proc of the International Conference on Crime Detection and Prevention. London, UK: IET, 2010. DOI: DOI:10.1049/ic.2009.0230. [本文引用:2]
WANG C, ZHANG J P, WANG L, et al. Human Identification Using Temporal Information Preserving Gait Template. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2012, 34(11): 2164-2176. [本文引用:2]
LIU J Y, ZHENG N N. Gait History Image: A Novel Temporal Template for Gait Recognition // Proc of the IEEE International Conference on Multimedia and Expo. Washington, USA: IEEE, 2007: 663-666. [本文引用:2]
GUAN Y, LI C T, HU Y J. Robust Clothing-Invariant Gait Recognition // Proc of the 8th IEEE International Conference on Intelligent Information Hiding and Multimedia Signal Processing. Wa-shington, USA: IEEE, 2012: 321-324. [本文引用:1]
HUANG S, ELGAMMAL A, LU J W, et al. Cross-Speed Gait Recognition Using Speed-Invariant Gait Templates and Globality-Locality Preserving Projections. IEEE Transactions on Information Forensics and Security, 2015, 10(10): 2071-2083. [本文引用:1]
SINGH S, BISWAS K K. Biometric Gait Recognition with Carrying and Clothing Variants // Proc of the 3rd International Conference on Pattern Recognition and Machine Intelligence. Berlin, Germany: Springer, 2009: 446-451. [本文引用:1]
ZHANG J P, PU J, CHEN C Y. Low-Resolution Gait Recognition. IEEE Transactions on Systems, Man, and Cybernetics(Cyberne-tics), 2010, 40(4): 986-996. [本文引用:1]
KUSAKUNNIRAN W, WU Q, ZHANG J, et al. Recognizing Gaits Across Views through Correlated Motion Co-clustering. IEEE Transactions on Image Processing, 2014, 23(2): 696-709. [本文引用:1]
BASHI K, XIANG T, GONG S G. Cross View Gait Recognition Using Correlation Strength // Proc of the British Machine Vision Conference. London, UK: BMVA, 2010: 1-11. [本文引用:1]
HU M D, WANG Y H, ZHANG Z X, et al. View-Invariant Discriminative Projection for Multi-view Gait-Based Human Identification. IEEE Transactions on Information Forensics and Security, 2013, 8(12): 2034-2045. [本文引用:1]
XING X L, WANG K J, YAN T, et al. Complete Canonical Correlation Analysis with Application to Multi-view Gait Recognition. Pattern Recognition, 2016, 50: 107-117. [本文引用:1]
MAKIHARA Y, SAGAWA R, MUKAIGAWA Y, et al. Gait Re-cognition Using a View Transformation Model in the Frequency Domain // Proc of the European Conference on Computer Vision. Berlin, Germany: Springer, 2006, III: 151-163. [本文引用:2]
KUSAKUNNIRAN W, WU Q, ZHANG J, et al. Support Vector Regression for Multi-view Gait Recognition Based on Local Motion Feature Selection // Proc of the IEEE Conference on Computer Vision and Pattern Recognition. Washington, USA: IEEE, 2010: 974-981. [本文引用:1]
KUSAKUNNIRAN W, WU Q, LI H D, et al. Multiple Views Gait Recognition Using View Transformation Model Based on Optimized Gait Energy Image // Proc of the 12th IEEE Conference on Computer Vision and Pattern Recognition. Washington, USA: IEEE, 2010: 1058-1064. [本文引用:1]
KRIZHEVSKY A, SUTSKEVER I, HINTON G E. ImageNet Cla-ssification with Deep Convolutional Neural Networks // Proc of the 26th Annual Conference on Neural Information Processing Systems. Cambridge, USA: The MIT Press, 2012: 1097-1105. [本文引用:1]
BENGIO Y, SIMARD P Y, FRASCONI P. Learning Long-Term Dependencies with Gradient Descent Is Difficult. IEEE Transactions on Neural Networks, 1994, 5(2): 157-166. [本文引用:3]
ZHANG X F, SUN S Q, LI C, et al. DeepGait: A Learning Deep Convolutional Representation for Gait Recognition // Proc of the Chinese Conference on Biometric Recognition. Berlin, Germany: Springer, 2017: 447-456. [本文引用:1]
SIMONYAN K, ZISSERMAN A. Very Deep Convolutional Networks for Large-Scale Image Recognition[C/OL]. [2017-11-25]. https://arxiv.org/pdf/1409.1556.pdf. [本文引用:1]
DENG J, DONG W, SOCHER R, et al. ImageNet: A Large-Scale Hierarchical Image Database // Proc of the IEEE Conference on Computer Vision and Pattern Recognition. Washington, USA: IEEE, 2009: 248-255. [本文引用:1]
SHIRAGA K, MAKIHARA Y, MURAMATSU D, et al. GEINet: View-Invariant Gait Recognition Using a Convolutional Neural Network // Proc of the International Conference on Biometrics. Wa-shington, USA: IEEE, 2016. DOI: DOI:10.1109/ICE.2016.7550060. [本文引用:2]
WOLF T, BABAEE M, RIGOLL G. Multi-view Gait Recognition Using 3D Convolutional Neural Networks // Proc of the IEEE International Conference on Image Processing. Washington, USA: IEEE, 2016: 4165-4169. [本文引用:3]
TRAN D, BOURDEV L, FERGUS R, et al. Learning Spatiotemporal Features with 3D Convolutional Networks // Proc of the IEEE International Conference on Computer Vision. Washington, USA: IEEE, 2015: 4489-4497. [本文引用:2]
SUN D Q, ROTH S, BLACK M J. Secrets of Optical Flow Estimation and Their Principles // Proc of the IEEE Conference on Computer Vision and Pattern Recognition. Washington, USA: IEEE, 2010: 2432-2439. [本文引用:1]
THAPAR D, AGGARWAL D, AGARWAL P, et al. VGR-Net: A View Invariant Gait Recognition Network[C/OL]. [2017-10-13]. https://arxiv.org/pdf/1710.04803.pdf. [本文引用:1]
HADSELL R, CHOPRA S, LECUN Y. Dimensionality Reduction by Learning an Invariant Mapping // Proc of the IEEE Conference on Computer Vision and Pattern Recognition. Washington, USA: IEEE, 2006: 1735-1742. [本文引用:1]
SCHROFF F, KALENICHENKO D, PHILBIN J. FaceNet: A Unified Embedding for Face Recognition and Clustering // Proc of the IEEE Conference on Computer Vision and Pattern Recognition. Washington, USA: IEEE, 2015: 815-823. [本文引用:2]
LIAO R J, CAO C S, GARCIA E B, et al. Pose-Based Temporal-Spatial Network(PTSN) for Gait Recognition with Carrying and Clothing Variations // Proc of the Chinese Conference on Biometric Recognition. Berlin, Germany: Springer, 2017: 474-483. [本文引用:5]
HOCHREITER S, SCHMIDHUBER J. Long Short-Term Memory. Neural Computation, 1997, 9(8): 1735-1780. [本文引用:1]
GRAVES A, MOHAMED A R, HINTON G. Speech Recognition with Deep Recurrent Neural Networks // Proc of the IEEE International Conference on Acoustics, Speech and Signal Processing. Washington, USA: IEEE, 2013: 6645-6649. [本文引用:1]
SUTSKEVER I, VINYALS O, LE Q V. Sequence to Sequence Learning with Neural Networks // Proc of the 28th Annual Confe-rence on Neural Information Processing Systems. Cambridge, USA: The MIT Press, 2014: 3104-3112. [本文引用:1]
WU Z F, HUANG Y Z, WANG L, et al. A Comprehensive Study on Cross-View Gait Based Human Identification with Deep CNNs. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 39(2): 209-226. [本文引用:5]
CHEN Q, WANG Y H, LIU Z, et al. Feature Map Pooling for Cross-View Gait Recognition Based on Silhouette Sequence Images[C/OL]. [2017-11-26]. https://arxiv.org/pdf/1711.0935-8.pdf. [本文引用:1]
FENG Y, LI Y C, LUO J B. Learning Effective Gait Features Using LSTM // Proc of the 23rd IEEE International Conference on Pattern Recognition. Washington, USA: IEEE, 2017: 325-330. [本文引用:3]
PFISTER T, CHARLES J, ZISSERMAN A. Flowing ConvNets for Human Pose Estimation in Videos // Proc of the IEEE International Conference on Computer Vision. Washington, USA: IEEE, 2016: 1913-1921. [本文引用:1]
YU S Q, CHEN H F, REYES E B G, et al. GaitGAN: Invariant Gait Feature Extraction Using Generative Adversarial Networks // Proc of the IEEE Conference on Computer Vision and Pattern Re-cognition. Washington, USA: IEEE, 2017: 30-37. [本文引用:4]
MIRZA M, OSINDERO S. Conditional Generative Adversarial Nets [C/OL]. [2014-11-06]. https://arxiv.org/pdf/1411.1784.pdf. [本文引用:1]
CHEN X, DUAN Y, HOUTHOOFT R, et al. InfoGAN: Interpre-table Representation Learning by Information Maximizing Generative Adversarial Nets // LEE D D, SUGIYAMA M, LUXBURG U V, et al. , eds. Advances in Neural Information Processing Systems 29. Cambridge, USA: The MIT Press, 2016: 2172-2180. [本文引用:1]
YOO D, KIM N, PAR S, et al. Pixel-Level Domain Transfer // Proc of the European Conference on Computer Vision. Berlin, Germany: Springer, 2016: 517-532. [本文引用:1]
YU S Q, CHEN H F, WANG Q, et al. Invariant Feature Extraction for Gait Recognition Using Only One Uniform Model. Neurocomputing, 2017, 239: 81-93. [本文引用:4]
ZHENG L, BIE Z, SUN Y F, et al. Mars: A Video Benchmark for Large-Scale Person Re-identification // Proc of the European Conference on Computer Vision. Berlin, Germany: Springer, 2016: 868-884. [本文引用:1]
LU J W, TAN Y P. Gait-Based Human Age Estimation. IEEE Transactions on Information Forensics and Security, 2010, 5(4): 761-770. [本文引用:1]
YU S Q, TAN T N, HUANG K Q, et al. A Study on Gait-Based Gender Classification. IEEE Transactions on Image Processing, 2009, 18(8): 1905-1910. [本文引用:1]
EBENEZER R H P I. Robust Analytics for Video-Based Gait Biometrics. Ph. D. Dissertation. Chennai, India: Anna University, 2017. [本文引用:1]
O'SULLIVAN J D, SAID C M, DILLON L C, et al. Gait Analysis in Patients with Parkinson's Disease and Motor Fluctuations: Influence of Levodopa and Comparison with Other Measures of Motor Function. Movement Disorders, 1998, 13(6): 900-906. [本文引用:1]
MURO-DE-LA-HERRAN A, GARCIA-ZAPIRAIN B, MENDEZ-ZORRILLA A. Gait Analysis Methods: An Overview of Wearable and Non-wearable Systems, Highlighting Clinical Applications. Sensors, 2014, 14(2): 3362-3394. [本文引用:1]