Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
153 tokens/sec
GPT-4o
7 tokens/sec
Gemini 2.5 Pro Pro
45 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
38 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

A Novel Dual-pooling Attention Module for UAV Vehicle Re-identification (2306.14104v1)

Published 25 Jun 2023 in cs.CV

Abstract: Vehicle re-identification (Re-ID) involves identifying the same vehicle captured by other cameras, given a vehicle image. It plays a crucial role in the development of safe cities and smart cities. With the rapid growth and implementation of unmanned aerial vehicles (UAVs) technology, vehicle Re-ID in UAV aerial photography scenes has garnered significant attention from researchers. However, due to the high altitude of UAVs, the shooting angle of vehicle images sometimes approximates vertical, resulting in fewer local features for Re-ID. Therefore, this paper proposes a novel dual-pooling attention (DpA) module, which achieves the extraction and enhancement of locally important information about vehicles from both channel and spatial dimensions by constructing two branches of channel-pooling attention (CpA) and spatial-pooling attention (SpA), and employing multiple pooling operations to enhance the attention to fine-grained information of vehicles. Specifically, the CpA module operates between the channels of the feature map and splices features by combining four pooling operations so that vehicle regions containing discriminative information are given greater attention. The SpA module uses the same pooling operations strategy to identify discriminative representations and merge vehicle features in image regions in a weighted manner. The feature information of both dimensions is finally fused and trained jointly using label smoothing cross-entropy loss and hard mining triplet loss, thus solving the problem of missing detail information due to the high height of UAV shots. The proposed method's effectiveness is demonstrated through extensive experiments on the UAV-based vehicle datasets VeRi-UAV and VRU.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (61)
  1. Zhu, J. et al. Vehicle re-identification using quadruple directional deep learning features. \JournalTitleIEEE Transactions on Intelligent Transportation Systems 21, 410–420 (2020).
  2. Exploring spatial significance via hybrid pyramidal graph network for vehicle re-identification. \JournalTitleIEEE Transactions on Intelligent Transportation Systems 23, 8793–8804 (2022).
  3. He, S. et al. Multi-domain learning and identity mining for vehicle re-identification. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 582–583 (2020).
  4. Vehiclenet: Learning robust visual representation for vehicle re-identification. \JournalTitleIEEE Transactions on Multimedia 23, 2683–2693 (2021).
  5. Rong, L. et al. A vehicle re-identification framework based on the improved multi-branch feature fusion network. \JournalTitleScientific Reports 11, 1–12 (2021).
  6. Triplet contrastive learning for unsupervised vehicle re-identification. \JournalTitlearXiv preprint arXiv:2301.09498 (2023).
  7. Applications of unmanned aerial vehicle (uav) in road safety, traffic and highway infrastructure management: Recent advances and challenges. \JournalTitleTransportation research part A: policy and practice 141, 116–129 (2020).
  8. Development of uav-based target tracking and recognition systems. \JournalTitleIEEE Transactions on Intelligent Transportation Systems 21, 3409–3422 (2020).
  9. Network in network. \JournalTitlearXiv preprint arXiv:1312.4400 (2013).
  10. Attention-aware generalized mean pooling for image retrieval. \JournalTitlearXiv preprint arXiv:1811.00202 (2018).
  11. Towards explaining anomalies: a deep taylor decomposition of one-class models. \JournalTitlePattern Recognition 101, 107198 (2020).
  12. Refining activation downsampling with softpool. In Proceedings of the IEEE/CVF international conference on computer vision, 10357–10366 (2021).
  13. Zhai, S. et al. S3pool: Pooling with stochastic spatial sampling. In Proceedings of the IEEE conference on computer vision and pattern recognition, 4970–4978 (2017).
  14. Learned-norm pooling for deep feedforward and recurrent neural networks. In Machine Learning and Knowledge Discovery in Databases: European Conference, ECML PKDD 2014, Nancy, France, September 15-19, 2014. Proceedings, Part I 14, 530–546 (Springer, 2014).
  15. Liftpool: Bidirectional convnet pooling. \JournalTitlearXiv preprint arXiv:2104.00996 (2021).
  16. Learning pooling for convolutional neural network. \JournalTitleNeurocomputing 224, 96–104 (2017).
  17. Ram: a region-aware deep model for vehicle re-identification. In 2018 IEEE International Conference on Multimedia and Expo (ICME), 1–6 (IEEE, 2018).
  18. Partition and reunion: A two-branch neural network for vehicle re-identification. In CVPR Workshops, 184–192 (2019).
  19. Wang, Z. et al. Orientation invariant feature embedding and spatial temporal regularization for vehicle re-identification. In Proceedings of the IEEE international conference on computer vision, 379–387 (2017).
  20. Zhang, G. et al. Sha-mtl: soft and hard attention multi-task learning for automated breast cancer ultrasound image segmentation and classification. \JournalTitleInternational Journal of Computer Assisted Radiology and Surgery 16, 1719–1725 (2021).
  21. Shen, F. et al. Hsgm: A hierarchical similarity graph module for object re-identification. In 2022 IEEE International Conference on Multimedia and Expo (ICME), 1–6 (IEEE, 2022).
  22. Pan, X. et al. Vehicle re-identification approach combining multiple attention mechanisms and style transfer. In 2022 3rd International Conference on Pattern Recognition and Machine Learning (PRML), 65–71 (IEEE, 2022).
  23. Bai, Y. et al. Group-sensitive triplet embedding for vehicle reidentification. \JournalTitleIEEE Transactions on Multimedia 20, 2385–2399 (2018).
  24. Vehicle and person re-identification with support neighbor loss. \JournalTitleIEEE Transactions on Neural Networks and Learning Systems 33, 826–838 (2022).
  25. Cross domain knowledge learning with dual-branch adversarial network for vehicle re-identification. \JournalTitleNeurocomputing 401, 133–144 (2020).
  26. Song, L. et al. Unsupervised domain adaptive re-identification: Theory and practice. \JournalTitlePattern Recognition 102, 107173 (2020).
  27. Vr-proud: Vehicle re-identification using progressive unsupervised deep architecture. \JournalTitlePattern Recognition 90, 52–65 (2019).
  28. Global reference attention network for vehicle re-identification. \JournalTitleApplied Intelligence 1–16 (2022).
  29. Wang, Q. et al. Viewpoint adaptation learning with cross-view distance metric for robust vehicle re-identification. \JournalTitleInformation Sciences 564, 71–84 (2021).
  30. Vehicle re-identification in still images: Application of semi-supervised learning and re-ranking. \JournalTitleSignal Processing: Image Communication 76, 261–271 (2019).
  31. Wang, Q. et al. Inter-domain adaptation label for data augmentation in vehicle re-identification. \JournalTitleIEEE Transactions on Multimedia 24, 1031–1041 (2022).
  32. Shen, F. et al. An efficient multiresolution network for vehicle reidentification. \JournalTitleIEEE Internet of Things Journal 9, 9049–9059 (2022).
  33. Git: Graph interactive transformer for vehicle re-identification. \JournalTitleIEEE Transactions on Image Processing 32, 1039–1051 (2023).
  34. Wang, P. et al. Vehicle re-identification in aerial imagery: Dataset and approach. In Proceedings of the IEEE/CVF International Conference on Computer Vision, 460–469 (2019).
  35. An adaptively attention-driven cascade part-based graph embedding framework for uav object re-identification. \JournalTitleRemote Sensing 14, 1436 (2022).
  36. Attention mask-based network with simple color annotation for uav vehicle re-identification. \JournalTitleIEEE Geoscience and Remote Sensing Letters 19, 1–5 (2022).
  37. Vehicle re-identification in aerial imagery based on normalized virtual softmax loss. \JournalTitleApplied Sciences 12, 4731 (2022).
  38. Enhancing part features via contrastive attention module for vehicle re-identification. In 2022 IEEE International Conference on Image Processing (ICIP), 1816–1820 (IEEE, 2022).
  39. Dual-relational attention network for vehicle re-identification. \JournalTitleApplied Intelligence 53, 7776–7787 (2023).
  40. Squeeze-and-excitation networks. In Proceedings of the IEEE conference on computer vision and pattern recognition, 7132–7141 (2018).
  41. Non-local neural networks. In Proceedings of the IEEE conference on computer vision and pattern recognition, 7794–7803 (2018).
  42. Rotate to attend: Convolutional triplet attention module. In Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, 3139–3148 (2021).
  43. Zhang, J. et al. Dual attention granularity network for vehicle re-identification. \JournalTitleNeural Computing and Applications 1–12 (2022).
  44. Hendrycks, D. et al. Augmix: A simple data processing method to improve robustness and uncertainty. \JournalTitlearXiv preprint arXiv:1912.02781 (2019).
  45. Rethinking the inception architecture for computer vision. In Proceedings of the IEEE conference on computer vision and pattern recognition, 2818–2826 (2016).
  46. Liu, C. et al. Posture calibration based cross-view & hard-sensitive metric learning for uav-based vehicle re-identification. \JournalTitleIEEE Transactions on Intelligent Transportation Systems 23, 19246–19257 (2022).
  47. Vehicle re-identification based on uav viewpoint: Dataset and method. \JournalTitleRemote Sensing 14, 4603 (2022).
  48. Large-scale vehicle re-identification in urban surveillance videos. In 2016 IEEE international conference on multimedia and expo (ICME), 1–6 (IEEE, 2016).
  49. Person re-identification by local maximal occurrence representation and metric learning. In Proceedings of the IEEE conference on computer vision and pattern recognition, 2197–2206 (2015).
  50. In defense of the triplet loss for person re-identification. \JournalTitlearXiv preprint arXiv:1703.07737 (2017).
  51. Chu, R. et al. Vehicle re-identification with viewpoint-aware metric learning. In Proceedings of the IEEE/CVF international conference on computer vision, 8282–8291 (2019).
  52. Yang, L. et al. Resolution adaptive networks for efficient inference. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2369–2378 (2020).
  53. Zhang, H. et al. Resnest: Split-attention networks. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2736–2746 (2022).
  54. Learning discriminative features with multiple granularities for person re-identification. In Proceedings of the 26th ACM international conference on Multimedia, 274–282 (2018).
  55. Scan: Spatial and channel attention network for vehicle re-identification. In Advances in Multimedia Information Processing–PCM 2018: 19th Pacific-Rim Conference on Multimedia, Hefei, China, September 21-22, 2018, Proceedings, Part III 19, 350–361 (Springer, 2018).
  56. He, L. et al. Fastreid: A pytorch toolbox for general instance re-identification. \JournalTitlearXiv preprint arXiv:2006.02631 (2020).
  57. Cbam: Convolutional block attention module. In Proceedings of the European conference on computer vision (ECCV), 3–19 (2018).
  58. Coordinate attention for efficient mobile network design. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 13713–13722 (2021).
  59. Sun, Y. et al. Circle loss: A unified perspective of pair similarity optimization. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 6398–6407 (2020).
  60. Multi-similarity loss with general pair weighting for deep metric learning. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 5022–5030 (2019).
  61. Khosla, P. et al. Supervised contrastive learning. \JournalTitleAdvances in neural information processing systems 33, 18661–18673 (2020).
Citations (4)
View on Semantic Scholar

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now