Abstract
Widely used surveillance cameras have promoted large amounts of street scene data, which contains one important but long-neglected object: the vehicle. Here we focus on the challenging problem of vehicle model verification. Most previous works usually employ global features (e.g., fully connected features) to further perform vehicle-level deep metric learning (e.g., triplet-based network). However, we argue that it is noteworthy to investigate the distinctiveness of local features and consider vehicle-part-level metric learning by reducing the intra-class variance as much as possible. In this article, we introduce a simple yet powerful deep model—the enforced intra-class alignment network (EIA-Net)—which can learn a more discriminative image representation by localizing key vehicle parts and jointly incorporating two distance metrics: vehicle-level embedding and vehicle-part-sensitive embedding. For learning features, we propose an effective feature extraction module that is composed of two components: the regional proposal network (RPN)-based network and part-based CNN. The RPN is used to define key vehicle regions and aggregate local features on these regions, whereas part-based CNN offers supplementary global features for the RPN-based network. The fusion features learned by feature extraction module are cast into the deep metric learning module. Especially, we derived an enforced intra-class alignment loss by re-utilizing key vehicle part information to enhance reducing intra-class variance. Furthermore, we modify the coupled cluster loss to model the vehicle-level embedding by enlarging the inter-class variance while shortening intra-class variance. Extensive experiments over benchmark datasets VehicleID and CompCars have shown that the proposed EIA-Net significantly outperforms the state-of-the-art approaches for vehicle model verification. Furthermore, we also conduct comprehensive experiments on vehicle re-identification datasets (i.e., VehicleID and VeRi776) to validate the generalization ability effectiveness of our proposed method.
- [1] 2019. Multi-label-based similarity learning for vehicle re-identification. IEEE Access 7, 16 (2019), 2605–2616.Google Scholar
- [2] 2018. Group-sensitive triplet embedding for vehicle re-identification. IEEE Trans. Multimedia 20, 9 (2018), 2385–2399. Google Scholar
Digital Library
- [3] 2012. Bayesian face revisited: A joint formulation. In Proc. Eur. Conf. Comp. Vis. Google Scholar
Digital Library
- [4] 2016. DeepLab: Semantic image segmentation with deep convolutional nets, atrous convolution, and fully connected CRFs. IEEE Trans. Pattern Anal. Mach. Intell. 21, 1 (2016), 1–13.Google Scholar
- [5] 2017. Beyond triplet loss: A deep quadruplet network for person re-identification.. In Proc. IEEE Conf. Comp. Vis. Patt. Recogn.Google Scholar
- [6] 2016. Person re-identification by multi-channel parts-based CNN with improved triplet loss function. In Proc. IEEE Conf. Comp. Vis. Patt. Recogn.Google Scholar
- [7] 2005. Learning a similarity metric discriminatively,with application to face verification. In Proc. IEEE Conf. Comp. Vis. Patt. Recogn. Google Scholar
Digital Library
- [8] 2016. Low resolution vehicle re-identification based on appearance features for wide area motion imagery. In Proc. IEEE Winter Appl. Comput. Vis. Workshops.Google Scholar
- [9] 2018. General knowledge embedded image representation learning. IEEE Trans. Multimedia 20, 1 (2018), 198–207. Google Scholar
Digital Library
- [10] 2015. Aggregating deep convolutional features for image retrieval. In Proc. IEEE Int. Conf. Comp. Vis.Google Scholar
- [11] 2012. Group-sensitive triplet embedding for vehicle re-identification. IEEE Trans. Multimedia 14, 1 (2012), 28–42. Google Scholar
Digital Library
- [12] 2016. Deep image retrieval: Learning global representations for image search. In Proc. Eur. Conf. Comp. Vis.Google Scholar
- [13] 2018. Learning coarse-to-fine structured feature embedding for vehicle re-identification. In Proc. AAAI Conf. Art. Intell. Google Scholar
Digital Library
- [14] 2019. Part-regularized near-duplicate vehicle re-identification. In Proc. IEEE Conf. Comp. Vis. Patt. Recogn.Google Scholar
- [15] 2016. Deep residual learning for image recognition. In Proc. IEEE Conf. Comp. Vis. Patt. Recogn.Google Scholar
- [16] 2014. From generic to specific deep representations for visual recognition. arXiv preprint arXiv:1406.5774v1 (2014).Google Scholar
- [17] 2014. Car make and model recognition using 3D curve alignment. In Proc. IEEE Winter Appl. Comput. Vis.Google Scholar
- [18] 2014. Discriminative deep metric learning for face verification in the wild. In Proc. IEEE Conf. Comp. Vis. Patt. Recogn. Google Scholar
Digital Library
- [19] 2019. A dual-path model with adaptive attention for vehicle re-identification. In Proc. IEEE Int. Conf. Comp. Vis.Google Scholar
- [20] 2013. 3D object representations for fine-grained categorization. In Proc. IEEE Int. Conf. Comp. Vis. Workshops. Google Scholar
Digital Library
- [21] 2012. ImageNet classification with deep convolutional neural networks. In Proc. Adv. Neural Inf. Process. Syst. Google Scholar
Digital Library
- [22] 2015. Weakly supervised deep metric learning for community-contributed image retrieval. IEEE Trans. Multimedia 17, 11 (2015), 1989–1999.Google Scholar
Digital Library
- [23] 2015. Person re-identification by local maximal occurrence representation and metric learning. In Proc. IEEE Conf. Comp. Vis. Patt. Recogn.Google Scholar
- [24] 2014. Jointly optimizing 3D model fitting and fine-grained classification. In Proc. Eur. Conf. Comp. Vis.Google Scholar
- [25] 2016. Deep relative distance learning: Tell the difference between similar vehicles. In Proc. IEEE Conf. Comp. Vis. Patt. Recogn.Google Scholar
- [26] 2016. A deep learning-based approach to progressive vehicle re-identification for urban surveillance. In Proc. Eur. Conf. Comp. Vis.Google Scholar
- [27] 2018. PROVID: Progressive and multimodal vehicle re-identification for large-scale urban surveillance. IEEE Trans. Multimedia 20, 3 (2018), 645–658. Google Scholar
Digital Library
- [28] 2020. Beyond the parts: Learning multi-view cross-part correlation for vehicle re-identification. In Proc. ACM Multimedia. Google Scholar
Digital Library
- [29] 2018. RAM: A region-aware deep model for vehicle re-identification. In Proc. IEEE Int. Conf. Multimedia Expo.Google Scholar
- [30] 2015. Fully convolutional networks for semantic segmentation. In Proc. IEEE Conf. Comp. Vis. Patt. Recogn.Google Scholar
- [31] 2019. Embedding adversarial learning for vehicle re-identification. IEEE Trans. Image Proc. 28, 8 (2019), 3794–3807.Google Scholar
Cross Ref
- [32] 2019. Veri-Wild: A large dataset and a new method for vehicle re-identification in the wild. In Proc. IEEE Conf. Comp. Vis. Patt. Recogn.Google Scholar
- [33] 2019. GCAN: Graph convolutional adversarial network for unsupervised domain adaptation. In Proc. IEEE Conf. Comp. Vis. Patt. Recogn.Google Scholar
- [34] 2020. Parsing-based view-aware embedding network for vehicle re-identification. In Proc. IEEE Conf. Comp. Vis. Patt. Recogn.Google Scholar
- [35] 2019. Stripe-based and attribute-aware network: A two-branch deep model for vehicle re-identification. arXiv preprint arXiv 1910.05549 (2019).Google Scholar
- [36] 2014. Car make and model recognition using 3D curve alignment. In Proc.IEEE Conf. Appl. Comp. Vis.Google Scholar
- [37] 2015. Faster R-CNN: Towards real-time object detection with region proposal networks. In Proc. Adv. Neural Inf. Process. Syst. Google Scholar
Digital Library
- [38] 2015. FaceNet: A unified embedding for face recognition and clustering. In Proc. IEEE Conf. Comp. Vis. Patt. Recogn.Google Scholar
- [39] 2017. Grad-CAM: Visual explanations from deep networks via gradient-based localization. arXiv preprint arXiv:1610.02391 (2017).Google Scholar
- [40] 2013. OverFeat: Integrated recognition, localization and detection using convolutional networks. arXiv preprint arXiv:1312.6229 (2013).Google Scholar
- [41] 2017. Learning deep neural networks for vehicle Re-ID with visual-spatio-temporalpath proposals. In Proc. IEEE Int. Conf. Comp. Vis.Google Scholar
- [42] 2015. Discriminative learning of deep convolutional feature point descriptors. In Proc. IEEE Int. Conf. Comp. Vis. Google Scholar
Digital Library
- [43] 2012. Very deep convolutional networks for large-scale image recognition. In ICLR.Google Scholar
- [44] 2015. Very deep convolutional networks for large-scale image recognition. In Proc. Int. Conf. Learn. Repr.Google Scholar
- [45] 2016. Deep metric learning via lifted structured feature embedding. In Proc. IEEE Conf. Comp. Vis. Patt. Recogn.Google Scholar
- [46] 2014. Deep learning face representation by joint identification-verification. In Proc. Adv. Neural Inf. Process. Syst. Google Scholar
Digital Library
- [47] 2014. Deep learning face representation from predicting 10,000 classes. In Proc. IEEE Conf. Comp. Vis. Patt. Recogn. Google Scholar
Digital Library
- [48] 2015. Deeply learned face representations are sparse, selective, and robust. In Proc. IEEE Conf. Comp. Vis. Patt. Recogn.Google Scholar
- [49] 2014. DeepFace: Closing the gap to human-level performance in face verification. In Proc. IEEE Conf. Comp. Vis. Patt. Recogn. Google Scholar
Digital Library
- [50] 2019. PAMTRI: Pose-aware multi-task learning for vehicle re-identification using highly randomized synthetic data. In Proc. IEEE Int. Conf. Comp. Vis.Google Scholar
- [51] 2016. Particular object retrieval with integral max-pooling of CNN activations. In Proc. Int. Conf. Learn. Repr.Google Scholar
- [52] 2014. Learning fine-grained image similarity with deep ranking. In Proc. IEEE Conf. Comp. Vis. Patt. Recogn. Google Scholar
Digital Library
- [53] 2017. Orientation invariant feature embedding and spatial temporal regularization for vehicle re-identification. In Proc. IEEE Int. Conf. Comp. Vis.Google Scholar
- [54] 2015. Distance metric learning for large margin nearest neighbor classification. J. Mach. Learn. Res. 10, 1 (2015), 207–224. Google Scholar
Digital Library
- [55] 2015. Data-driven 3D voxel patterns for object category recognition. In Proc. IEEE Conf. Comp. Vis. Patt. Recogn.Google Scholar
- [56] 2018. Learning semantic representations for unsupervised domain adaptation. In Proc. Int. Conf. Mach. Learn.Google Scholar
- [57] 2018. Person re-identification with metric learning using privileged information. IEEE Trans. Image Proc. 27, 2 (2018), 791–805.Google Scholar
Cross Ref
- [58] 2015. A large-scale car dataset for fine-grained categorization and verification. In Proc. IEEE Conf. Comp. Vis. Patt. Recogn.Google Scholar
- [59] 2017. Enhancing person re-identification in a self-trained subspace. ACM Trans. Multimed. Comput. Commun. Appl. 13, 3 (2017), 1–23. Google Scholar
Digital Library
- [60] . 2014. Deep metric learning for person re-identification. In Proc. IEEE Int. Conf. Patt. Recogn. Google Scholar
Digital Library
- [61] 2016. Embedding label structures for fine-grained feature representation. In Proc. IEEE Conf. Comp. Vis. Patt. Recogn.Google Scholar
- [62] 2012. Three dimensional deformable-model-based localization and recognition of road vehicles. IEEE Trans. Image Proc. 21, 1 (2012), 1–13. Google Scholar
Digital Library
- [63] 2015. Scalable person re-identification: A benchmark. In Proc. IEEE Int. Conf. Comp. Vis. Google Scholar
Digital Library
- [64] 2020. VehicleNet: Learning robust visual representation for vehicle re-identification. IEEE Trans. Multimedia 23 (2020), 2683–2693.Google Scholar
- [65] 2016. Fine-grained image classification by exploring bipartite-graph labels. In Proc. IEEE Conf. Comp. Vis. Patt. Recogn.Google Scholar
- [66] 2017. Cross-view GAN based vehicle generation for re-identification. In Proc. Brit. Mach. Vis. Conf.Google Scholar
- [67] 2018. Aware attentive multi-view inference for vehicle re-identification. In Proc. IEEE Conf. Comp. Vis. Patt. Recogn.Google Scholar
- [68] 2018. Viewpoint-aware attentive multi-view inference for vehicle re-identification. In Proc. IEEE Conf. Comp. Vis. Patt. Recogn.Google Scholar
- [69] 2016. DeepM: A deep part-based model for object detection and semantic part localization. In Proc. Int. Conf. Learn. Repr.Google Scholar
Index Terms
Seeing Crucial Parts: Vehicle Model Verification via a Discriminative Representation Model
Recommendations
Discriminative-region attention and orthogonal-view generation model for vehicle re-identification
AbstractVehicle re-identification (Re-ID) is urgently demanded to alleviate the pressure caused by the increasingly onerous task of urban traffic management. Multiple challenges hamper the applications of vision-based vehicle Re-ID methods: (1) The ...
Joint Pyramid Feature Representation Network for Vehicle Re-identification
AbstractVehicle re-identification (Re-ID) technology plays an important role in the intelligent transportation system for smart city. Due to various uncertain factors in the real-world scenarios, (e.g., resolution variation, viewpoint variation, ...
Multi-Task Deep Metric Learning with Boundary Discriminative Information for Cross-Age Face Verification
AbstractImage based face verification has attracted extension attention in the fields of pattern recognition and intelligent vision. With difference in age, cross-age face verification from facial images remains a challenging work because of a large ...






Comments