发表机构
Beijing Normal-Hong Kong Baptist University; Hong Kong Baptist University; Jilin University; Guangdong Provincial Key Laboratory of Interdisciplinary Research and Application for Data Science; Nanyang Technological University（北京师范大学-香港浸会大学联合国际学院; 香港浸会大学; 吉林大学; 广东省数据科学交叉研究与应用重点实验室; 南洋理工大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
针对流式3D重建中循环场景状态易被噪声观测破坏的问题，提出可靠性校准学习率方法ReCal3R，在不增加额外开销的前提下大幅提升长序列重建精度。
AI 中文摘要
流式3D重建依赖紧凑的循环场景状态，以线性时间和有界内存处理长图像流。但重复更新会逐步破坏该状态，导致可靠历史信息被噪声或模糊观测覆盖。本文提出面向循环3D重建的可靠性校准学习率方法ReCal3R，该方法不直接使用候选学习率，而是从维护的场景状态中估计状态token的可靠性，以此校准由token对齐度、状态重建残差和近期更新压力推导得到的候选学习率。最终得到的逐token学习率在保守基准率和候选率之间插值，既抑制对不可靠token的激进更新，又保留对有效帧的适配能力。将其作为无训练校准规则应用于CUT3R后，ReCal3R在长序列的位姿、深度及重建质量上取得优异表现，绝对轨迹误差ATE降低3.7倍，同时保持相当的运行时间和内存占用。代码可访问：此https URL。
英文摘要
Streaming 3D reconstruction relies on a compact recurrent scene state to process long image streams in linear time and bounded memory. However, repeated updates can gradually corrupt this state, causing reliable historical information to be overwritten by noisy or ambiguous observations. We introduce ReCal3R, a reliability-calibrated learning rate method for recurrent 3D reconstruction. Instead of directly applying a candidate learning rate, our method estimates state token reliability from the maintained scene state and uses it to calibrate a candidate learning rate derived from token alignment, state reconstruction residual, and recent update pressure. The resulting token-wise learning rate interpolates between a conservative base rate and the candidate rate, suppressing aggressive updates on unreliable tokens while preserving adaptation to informative frames. Applied to CUT3R as a training-free calibration rule, ReCal3R reaches strong performance on long sequences in pose, depth, and reconstruction quality, including a 3.7$\times$ reduction in ATE, with comparable runtime and memory. Code is available at: https://github.com/Powertony102/ReCal3R.
Comments23 pages, 7 figures. Project Page: https://powertony102.github.io/recal3r.github.io/
