计算机科学 ›› 2024, Vol. 51 ›› Issue (11): 166-173.doi: 10.11896/jsjkx.230900078
范晓鹏, 彭力, 杨杰龙
FAN Xiaopeng, PENG Li, YANG Jielong
摘要： 目标跟踪是计算机视觉领域的一个基础研究问题。随着跟踪技术的发展,现存的跟踪器主要存在两个挑战,即依赖于大量的数据标注信息和跟踪漂移,它们严重限制了跟踪器性能的提升。为了应对以上挑战,提出了无监督目标跟踪和隐空间匹配的方法。首先,通过可校正光流方法在前景中生成图像对;其次,利用生成的图像对从头开始训练孪生跟踪器;最后,使用隐空间匹配的方法,解决了跟踪器在目标形变较大、遮挡、出视野和漂移等情况下跟丢的问题。实验结果表明,算法UHOT的性能在多个数据集上有显著提升,在困难场景下展现出了较强的鲁棒性。与最新的无监督算法SiamDF相比,UHOT在VOT 数据集上取得了8% 的增益,与最新的监督孪生跟踪器相当。
中图分类号:
| [1] WANG N,SONG Y B,MA C,et al.Unsupervised deep tracking[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.2019:1308-1317. [2] CHONG H S,MA Y J,CHEN J C,et al.S2siamfc:Self-supervised fully convolutional siamese network for visual tracking[C]//Proceedings of the 28th ACM International Conference on Multimedia.2020:1948-1957. [3] LIU L,ZHANG J N,HE R F,et al.Learning by analogy:Reliable supervision from transformations for unsupervised optical flow estimation[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.2020:6489-6498. [4] LUCA B,JACK V,JOAO F,et al.Fully-convolutional siamese networks for object tracking[C]//Computer Vision－ECCV 2016 Workshops:Amsterdam,The Netherlands,October 8－10 and 15－16,2016,Proceedings,Part II 14.Springer,2016:850-865. [5] CAI H Y,LAN L,ZHANG J,et al.Siamdf:Tracking training data-free siamese tracker[M].Neural Networks,2023. [6] LI H J,PENG L.High-speed tracking algorithm based on negative sample mining and feature fusion[J].Control and Decision Making,2023,38(9):2554-2562. [7] SUN K W,WANG Z H,LIU H,et al.Maximum Overlap Single Target Tracking Algorithm Based on Attention Mechanism.[J].Computer Science,2023,50(S1):397-401. [8] ZENG Z H,LUO H L.Cross-dataset Learning CombiningMulti-object Tracking and Human Pose Estimation[J].Compu-ter Science,2023,50(S1):512-518. [9] ZHENG J L,MA C,PENG H W,et al.Learning to track objects from unlabeled videos[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision.2021:13546-13555. [10] WU Q Q,WAN J,ANTONI B C.Progressive unsupervisedlearning for visual object tracking[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.2021:2993-3002. [11] ZHANG J,LIU Y,LIU H,et al.Learning local-global multiple correlation filters for robust visual tracking with Kalman filter redetection[J].Sensors,2021,21(4):1129. [12] LE N,RATHOUR V S,YAMAZAKI K,et al.Deep reinforcement learning in computer vision:a comprehensive survey[J].Artificial Intelligence Review,2022,55(4):2733-2819. [13] WANG Q,LI Z,LUCA B,et al.Fast online object tracking and segmentation:A unifying approach[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.2019:1328-1338. [14] CHENG Y M,LI L L,XU Y Y,et al.Segment and track anything[J].arXiv:2305.06558,2023. [15] LI B,YAN J J,WU W,et al.High performance visual tracking with siamese region proposal network[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.2018:8971-8980. [16] TARG S S,DIOGO A,KEVIN L.Resnet in resnet:Generalizing residual architectures[J].arXiv:1603.08029,2016. [17] YU J H,JIANG Y N,WANG Z Y,et al.Unitbox:An advanced object detection network[C]//Proceedings of the 24th ACM International Conference on Multimedia.2016:516-520. [18] HO Y S,SAMUEL W.The real-worldweight cross-entropy loss function:Modeling the costs of mislabeling[J].IEEE Access,2019,8:4806-4813. [19] JADERBERG M,SIMON K,ZISSERMAN A,et al.Spatialtransformer networks.[J].arXiv:1506.02025,2015. [20] JOOST V A,ANITHA K,MARC A R,et al.Transformation-based models of video sequences[J].arXiv:1701.08435,2017. [21] TOMAS J,ANKUSH G,HAKAN B,et al.Unsupervised lear-ning of object landmarks through conditional image generation[C]//Advances in Neural Information Processing Systems 31.2018. [22] OLAF R,PHILIPP F,THOMAS B.U-net:Convolutional net-works for biomedical image segmentation[C]//Medical Image Computing and Computer-Assisted Intervention-MICCAI 2015:18th Interna＝tional Conference,Munich,Germany,October 5－9,2015,Proceedings,Part III 18.Springer,2015:234-241. [23] ALIAK S,STEPHANE L,SERGEY T,et al.First order motion model for image animation[C]//Advances in Neural Information Processing Systems 32.2019. [24] MATEJ K,ALES L,JIRI M,et al.The sixth visual object tra-cking vot2018 challenge results[C]//Proceedings of theEuro-pean Conference on Computer Vision(ECCV).2018. [25] WU Y,LI J W,YANG M H.Online object tracking:A bench-mark[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.2013:2411-2418. [26] ZHANG Z P,PENG H W.Deeper and wider siamese networks for real-time visual tracking[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.2019:4591-4600. [27] LI Y M,FU C H,DING F Q,et al.Autotrack:Towards high-performance visual tracking for uav with automatic spatio-temporal regularization[C]//Proceedings of the IEEE/CVF Confe-rence on Computer Vision and Pattern Recognition(CVPR).2020. [28] MARTIN D,GOUTAM B,FAHAD S K,et al.Atom:Accuratetracking by overlap maximization[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.2019:4660-4669. |
|
||
首页