slam

HiP3R 提出一种层级式位姿图方法，通过先优化块级 Sim(3) 图建立全局布局，再固定该布局估计每帧共享的 SE(3) 相机位姿，将局部视频重建合并为全局一致的相机轨迹和 3D 几何，无需重新训练。

EgoXperience 提出一个统一框架，通过自动标注的大规模数据和单次前向传播模型，从第一人称视频中联合重建相机运动、场景几何、手部与物体，并发布了 EgoX-Data-6M 数据集与 EgoX-Bench 基准，取得显著优于基线的性能。

Maniformer 的新论文提出 MEgoVista，一套基于头戴式装置的物理 AI 数据系统，通过立体视觉和全序列优化，将手部手指误差降至 4.29 毫米、头部跟踪误差控制在 0.65–2.83 毫米，并经 26 相机动作捕捉实验室验证，为机器人学习提供可测量的人类动作数据。

Valerant is a training-free framework that transforms pretrained action-conditioned world models into world action models for exploring and constructing 3D game maps from a single image, reducing manual effort in game development.

This article lists five DIY robot projects, including an arm, a rover, a self-balancing robot, an ACT policy-trained robot, and a line follower, with details on parts, calibration, and tuning techniques.

TAMP-Nav is a unified framework that enhances embodied navigation by aligning vision-language models with 2D visual prompting, selective reasoning with compressed memory, and policy optimization, achieving state-of-the-art performance with high runtime and sample efficiency.

HiFi-UMI introduces a portable data-production system for robot-free UMI data that achieves high trajectory accuracy using stereo-inertial SLAM and wide-angle cameras. Training manipulation policies on this data alone enables zero-shot deployment on real robots, matching or exceeding teleoperation baselines across several model families, and the authors open-source a 2,000-hour high-fidelity dataset.

A tweet shares a free MIT guide called 'SLAM for Dummies' that teaches simultaneous localization and mapping for robotics, with code examples, and promotes a weekly robotics newsletter.

Grabette is an open, low-cost system for recording robot manipulation data using a handheld gripper and camera, aiming to build a shared dataset for robot learning.

This paper benchmarks six SLAM systems under low-light conditions using five LaMARia sequences, finding that only Kimera-VIO tracks all sequences to completion while other RGB-only systems fail or diverge, highlighting the need for inertial fusion and global optimization.

This paper presents an offline reinforcement learning framework for optimizing SLAM throughput control in warehouse fulfillment environments, balancing throughput maximization with downstream stability. The approach is algorithm-agnostic and demonstrates that the CQL policy improves system health by 22.97% and reduces throttling duration by 3.18%.

A visual breakdown of 8 major AI model architectures including LLMs, VLMs, MoE, SLMs, and more, plus a bonus mention of recursive language models from MIT.

This paper proposes Geodesic Flow Matching, a Riemannian transport method for denoising Spatial Semantic Pointers (SSPs) on toroidal manifolds, and demonstrates a 72% reduction in tracking error and 40% efficiency gain in a spiking neural SLAM system.

This week awesome-autoresearch added three items, including the autoslam project that applies Karpathy's autoresearch loop to LiDAR SLAM, and two blog posts analyzing the original experiments and revealing metric gaming.

Introduces LingBot-Map, a feed-forward 3D foundation model for streaming 3D reconstruction using a geometric context transformer architecture that achieves stable real-time performance at 20 FPS.