TM-APR：基于解析在线自适应的热时序-记忆定位
TM-APR: Thermal Temporal-Memory Localization via Analytic Online Adaptation
浏览论文内容
中文总结 AI 辅助
针对热视觉地点识别在线部署中的环境依赖与非线性漂移问题，本文首次将解析类增量学习与控制理论结合，提出TM-APR框架，通过无迹传播、高斯混合划分和H∞优化实现O(1)闭式更新，确保实时SLAM中更新延迟低于采集间隔。
中文摘要 AI 辅助
热视觉地点识别（Thermal VPR）将相机观测映射到已建图环境中的度量位姿，是自主导航的先决条件。然而，热VPR存在严重的环境依赖性、繁重的在线重训练开销，以及无法建模动态非线性偏移的问题，导致现有框架在在线部署时失效。为实现鲁棒的域不变地点识别，我们首次将解析类增量学习（ACIL）与域不变VPR相结合，揭示其无梯度矩阵更新构建了一个出人意料的强基线，性能优于传统微调。然而，标准ACIL因其结构线性假设，对极端非线性热波动表现出关键脆弱性。为克服此局限，我们利用ACIL与现代控制理论之间新颖的代数等价性，提出一个框架，将无迹传播（U-ACIL）、高斯混合划分（GMM-ACIL）和极小极大$H_\infty$优化（$H_\infty$-ACIL）直接嵌入更新循环。我们的公式保证了在$\mathcal{O}(1)$计算复杂度内的精确闭式矩阵更新，绕开反向传播，确保在线更新延迟（$\Delta t_{\mathrm{learn}}$）严格低于传感器采集间隔（$\Delta t_{\mathrm{acquire}}$），从而消除实时SLAM流水线中的轨迹跳变。
英文摘要
Thermal Visual Place Recognition (Thermal VPR) maps camera observations to metric poses within a mapped environment, serving as a prerequisite for autonomous navigation. However, thermal VPR suffers from severe environmental dependence, heavy online retraining overheads, and an inability to model dynamic non-linear shifts, causing existing frameworks to fail during online deployment. To achieve robust domain-invariant place recognition, we bridge Analytic Class-Incremental Learning (ACIL) with domain-invariant VPR for the first time, revealing that its gradient-free matrix updates construct a surprisingly strong baseline that outperforms conventional fine-tuning. Nevertheless, standard ACIL exhibits a critical vulnerability to extreme non-linear thermal fluctuations due to its structural linear assumptions. To overcome this limitation, we exploit a novel algebraic equivalence between ACIL and modern control theory, proposing a framework which embeds Unscented propagation (U-ACIL), Gaussian Mixture partitioning (GMM-ACIL), and minimax $H_\infty$ optimization ($H_\infty$-ACIL) directly into the update loop. Our formulation guarantees exact closed-form matrix updates within $\mathcal{O}(1)$ computational complexity, bypassing backpropagation to ensure that the online update latency ($Δt_{\mathrm{learn}}$) remains strictly bounded below the sensor acquisition interval ($Δt_{\mathrm{acquire}}$), thereby eliminating trajectory jumps in real-time SLAM pipelines.
