Trajectory-Anchor Optimization for Overconfident Thermal Visual Place Recognition: Zero-Leakage OOD Auditing and Kidnapped-Robot Recovery
作者: Zhiyuan Lu, Kanji Tanaka
分类: cs.RO, cs.CV
发布日期: 2026-07-06
备注: 11 pages, 5 figures, technical report
💡 一句话要点
提出轨迹锚优化以解决过度自信的热视觉位置识别问题
🎯 匹配领域: 支柱七:动作重定向 (Motion Retargeting) 支柱九:具身大模型 (Embodied Foundation Models)
关键词: 热视觉识别 轨迹锚优化 多视角验证 机器人导航 实时计算 分布外检测 几何一致性
📋 核心要点
- 现有的热视觉位置识别方法在处理分布外数据时容易产生过度自信的匹配错误,无法有效区分真实与虚假候选项。
- 论文提出的轨迹锚优化(TAO)通过将多视角验证转化为批量对齐问题,显著提高了计算效率,满足实时机器人需求。
- 实验结果表明,TAO在5米半径内有效过滤虚假候选,且在10米范围内实现了多视角几何一致性,显著降低了错误接受率。
📝 摘要(中文)
现代基于基础模型的热视觉位置识别(TIR-VPR)前端在闭集检索中表现出色,但在分布外(OOD)或未映射条件下,容易产生过度自信的强制匹配失败模式。它们在相似度评分不下降的情况下生成高度可信但错误的循环候选项。为了解决这一问题,本文提出了轨迹锚优化(TAO),通过将多视角时间验证压缩为批量SE(2) Procrustes对齐问题,显著降低了计算复杂度。TAO在严格的零泄漏评估协议下,能够有效隔离宏观尺度上的灾难性拓扑断裂,抑制关键的错误接受。
🔬 方法详解
问题定义:本文旨在解决现代热视觉位置识别系统在分布外或未映射条件下的过度自信匹配问题。现有方法在处理局部视觉模糊时,无法有效区分真实的度量定位误差与虚假候选项。
核心思路:论文提出的轨迹锚优化(TAO)通过将多视角时间验证压缩为批量SE(2) Procrustes对齐问题,利用张量级向量化和单次调用的批量SVD,避免了传统多假设跟踪(MHT)方法的动态树扩展,从而提高了计算效率。
技术框架:TAO的整体架构包括数据输入、轨迹锚生成、批量对齐处理和结果输出四个主要模块。首先,输入的多视角数据被处理以生成初步的轨迹锚,然后通过SE(2) Procrustes对齐进行批量验证,最后输出经过优化的匹配结果。
关键创新:TAO的主要创新在于其将多视角验证问题转化为批量对齐问题的能力,这一设计显著降低了计算复杂度,使得每帧的执行时间严格限制在O(KN)内,满足实时性要求。
关键设计:在TAO中,采用了批量SVD进行高效的矩阵分解,确保了在处理K=100个假设时的计算效率。此外,设计了严格的零泄漏评估协议,以确保在宏观尺度上有效隔离虚假候选。
🖼️ 关键图片
📊 实验亮点
实验结果显示,TAO在5米半径内有效过滤虚假候选,且在10米范围内实现了多视角几何一致性,显著降低了错误接受率。与传统方法相比,TAO在处理K=100个假设时,计算复杂度降低至O(KN),满足实时性需求。
🎯 应用场景
该研究的潜在应用领域包括自主机器人导航、无人驾驶汽车和智能监控系统等。通过提高热视觉位置识别的准确性和实时性,TAO能够在复杂环境中更好地支持机器人自主决策,提升其在动态场景中的适应能力和可靠性。
📄 摘要(原文)
Modern thermal visual place recognition (TIR-VPR) frontends based on foundation models achieve remarkable closed-set retrieval but suffer from an overconfident forced-matching failure mode. Under out-of-distribution (OOD) or unmapped conditions, they generate highly plausible yet false loop candidates without a drop in similarity scores. While classical multi-hypothesis tracking (MHT) backends can mitigate these ambiguities by maintaining divergent trajectory beliefs, their exponential computational overhead violates real-time robotic constraints. To bridge this gap, we present Trajectory-Anchor Optimization (TAO). To counter the combinatorial challenge of evaluating parallel hypotheses (e.g., K=100), TAO compresses multi-view temporal verification into a batched SE(2) Procrustes alignment problem. By leveraging tensor-level vectorization and single-invocation batched SVD, this formulation bypasses the dynamic tree expansion of MHT, guaranteeing a strictly bounded per-frame execution loop of O(KN). Under a strict zero-leakage evaluation protocol, we show that while a passive geometric backend cannot mathematically separate metric localization errors from coherent hallucinations at a micro-scale (<5m) due to local visual ambiguities, TAO serves as an efficient fail-safe filter at a macro-scale. Within a 5m radius, hallucinations often possess a locally consistent geometry that deceives rigid alignment. However, beyond this threshold, the K=100 disparate hypotheses disperse spatially across the global map. This dispersion breaks the rigid temporal co-visibility constraint within the sliding window (N=20), causing the joint optimization residual to escalate sharply. Consequently, TAO establishes a distinct macroscopic convergence basin (10m) where multi-view geometric consistency reliably isolates catastrophic topological breaks and suppresses critical false acceptances.