Considering tracking different types of multiple targets, the current tracking system's fixed model parameters fail to take full advantage of motion characteristics from different types of targets. As a result, the tracking accuracy may decrease and it even leads to track loss or mis-tracking. To address this issue, a Transformer-based intelligent data association and tracking method is put forward for different types of multiple targets. In the case of time sliding window, by integrating sparse attention and self-attention mechanisms, the target-type features and track association features are deeply explored and exploited from the historical radar data and short-time track data, to adaptively adjusting the process noise covariance matrix. This is beneficial to realize the refined motion modeling of different types of targets and acquire the reliable association between every radar echo and each target track. And it ensures to obtain high-quality tracking outputs for different types of multiple targets. Simulation results in the typical scenario of tracking dif-ferent types of multiple targets with varying clutter densities demonstrate that, compared to the classical joint probability data association filtering and intelligent data association and tracking method with fixed model parameters, the pro-posed method achieves superior tracking accuracy and stability.
[1]柳超, 王月基.对海探测雷达多目标跟踪技术综述[J].雷达学报, 2021, 10(1):100-115
[2]RANGESH A, Trivedi M M.No blind spots: Full-surround multi-object tracking for autonomous vehicles using cameras and lidars[J].IEEE Transactions on Intelligent Vehicles, 2019, 4(4):588-599
[3]ZHOU M, ZHANG J J, PAPANDREOU-SUPPAPPOLA A.Multiple target tracking in urban environments[J].IEEE Transactions on Signal Processing, 2015, 64(5):1270-1279
[4]FORTMANN T, BAR-SHALOM Y, SCHEFFE M.Sonar tracking of multiple targets using joint probabilis-tic data association[J].IEEE Journal of Oceanic Engineering, 1983, 8(3):173-184
[5]王树亮, 毕大平, 阮怀林.认知雷达波形自适应数据关联跟踪算法[J].宇航学报, 2017, 38(12):1331-1338
[6]Chong C Y.An overview of machine learning methods for multiple target tracking[J].2021 IEEE 24th Interna-tional Conference on Information Fusion (FUSION), Sun City, South Africa, 2021, :-
[7]LI X R.Tracking in clutter with strongest neighbor measurements[J].IEEE Transactions on Automatic Con-trol, 1998, 43(11):1560-1578
[8]SONG T L, LEE D G.A probabilistic nearest neighbor filter algorithm for m validated measurements[J].IEEE Transactions on Signal Processing, 2006, 54(7):2797-2802
[9]何友, 修建娟, 刘瑜, 等.雷达数据处理及应用(第四版)[J].北京: 电子工业出版社, 2022, :-
[10]WILLIAMS J L.Marginal multi-Bernoulli filters: RFS derivation of MHT, JIPDA, and association-based MeMBer[J].IEEE Transactions on Aerospace and Elec-tronic Systems, 2015, 51(3):1664-1687
[11]WILLETT P, RUAN Y, STREIT R.PMHT: Problems and some solutions[J].IEEE Transactions on Aerospace and Electronic Systems, 2002, 38(3):738-754
[12]SONG T L, LEE D G, RYU J.A probabilistic nearest neighbor filter algorithm for tracking in a clutter envi-ronment[J].Signal Processing, 2005, 85(10):2044-2053
[13]BAR-SHALOM Y, TSE E.Tracking in a cluttered envi-ronment with probabilistic data association[J].Automatica, 1975, 11(5):451-460
[14]He S M, SHIN H S, TSOURDOS A.Distributed joint probabilistic data association filter with hybrid fusion strategy[J].IEEE Transactions on Instrumentation and Measurement, 2019, 69(1):286-300
[15]QIAN H, YANG X, ZHANG X, et al.PLA-JPDA for indoor multi-person tracking using IR-UWB radars[J].IEEE Radar Conference, Florence, Italy, 2020, :-
[16]杨峰, 王永齐, 梁彦, 潘泉.基于概率假设密度滤波方法的多目标跟踪技术综述[J].自动化学报, 2013, 39(11):1944-1956
[17]申屠晗, 薛安克, 周治利.多传感器高斯混合融合多目标跟踪方法[J].自动化学报, 2017, 43(6):1028-1037
[18]徐洋, 徐晖, 安玮, 等.一种利用信号幅度信息的光学传感器像平面-滤波算法[J].宇航学报, 2012, 33(4):478-486
[19]SONNTAG V, LE CAILLEC J M, PERES A, et al.Transformer-based state estimation for tracking: Maneu-vering target and multi-target capabilities[J].2024 IEEE Radar Conference (RadarConf24), Denver, Colorado, USA, 2024, :-
[20]ZHANG Y, LI G, ZHANG X P, et al.A deep learning model based on transformer structure for radar tracking of maneuvering targets[J].Information Fusion, 2024, 103(无):102120-1-102120-12
[21]CHUNG Y N, CHOU P H, YANG M R, et al.Multiple-target tracking with competitive Hopfield neural network based data association[J].IEEE Transactions on Aero-space and Electronic Systems, 2007, 43(3):1180-1188
[22]LIU H J, ZHANG H, and MERTZ C.DeepDA: LSTM-based deep data association network for multitargets tracking in clutter[J].2019 22th International Confer-ence on Information Fusion (FUSION), Ottawa, Canada, 2019, :-
[23]MEINHARDT T, KIRILLOV A, LEAL-TAIXE L, et al.Trackformer: Multi-object tracking with transformers[J].IEEE/CVF Conference on Computer Vision and Pattern Recognition, New Orleans, Louisiana, USA, 2022, :-
[24]李文娜, 张顺生, 王文钦.基于网络的机载雷达多目标跟踪方法[J].雷达学报, 2022, 11(3):469-478
[25]PINTO J, HESS G, LJUNGBERGH W, et al.Next generation multitarget trackers: Random finite set methods vs transformer-based deep learning[J].2021 IEEE 24th International Conference on Information Fusion (FUSION), Sun City, South Africa, 2021, :-
[26]WEI X W, LIN Y R, ZHANG L N, et al.Transformer-based Multi-Target Tracking with Bayesian Perspective[J].2024 27th International Conference on Information Fu-sion (FUSION), Venice, Italy, 2024, :-
CJA