发表机构
Hangzhou Applied Acoustics Research Institute（杭州应用声学研究所）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
针对Transformer多目标跟踪的冗余计算问题，提出因果神经集合滤波（CNSF），通过仅编码当前测量并递归传递状态，结合Sinkhorn关联和卡尔曼更新，在模拟测试集上显著降低GOSPA并提升推理速度。
AI 中文摘要
基于Transformer的多目标跟踪（MTT）联合学习数据关联和状态估计，但MT3/Track-MT3风格的跟踪器反复重新编码测量窗口，导致冗余计算。我们提出因果神经集合滤波（CNSF），一种神经集合滤波器，仅编码当前测量，同时将过去的证据携带在结构化的递归跟踪状态中。CNSF结合了排他性Sinkhorn关联、带矩匹配的关联条件卡尔曼形更新以及带测量驱动出生的递归伯努利生命周期建模。这些机制施加软一对一约束，传播关联引起的状态不确定性，并支持在漏检和出生-死亡转换下的存在性估计。在保留的三机制模拟测试集上，CNSF相对于Track-MT3将平均GOSPA和T-GOSPA分别降低了19.3%和30.4%，参数减少55.9%，单线程CPU推理速度提升3.76倍。
英文摘要
Transformer-based multi-target tracking (MTT) jointly learns data association and state estimation, but MT3/Track-MT3-style trackers repeatedly re-encode measurement windows, incurring redundant computation. We propose Causal Neural Set Filtering (CNSF)\footnote{\href{https://github.com/daihuangyu/CNSF}{Code: https://github.com/daihuangyu/CNSF}}, a neural set filter that encodes only current measurements while carrying past evidence in a structured recursive track state. CNSF combines exclusive Sinkhorn association, association-conditioned Kalman-shaped updates with moment matching, and recurrent Bernoulli lifecycle modeling with measurement-driven birth. These mechanisms impose soft one-to-one constraints, propagate association-induced state uncertainty, and support existence estimation under missed detections and birth--death transitions. On a held-out three-regime simulated test set, CNSF reduces mean GOSPA and T-GOSPA relative to Track-MT3 by 19.3\% and 30.4\%, with 55.9\% fewer parameters and a $3.76\times$ speedup in single-thread CPU inference.
Comments5 pages, 2 figures
