The paper introduces a compact visual navigation system that decomposes the task into three analytically‑computed geometric interfaces and three small learned modules: an egress predictor, a navigation predictor, and an endpoint‑pinned residual diffusion generator. Only 0.58 M of the 23 M parameters are trained on 44 k frames, achieving competitive success rates and the lowest collision rate among evaluated methods across 6 060 point‑goal episodes in 60 environments. The design allows further parameter reduction by replacing the frozen image encoder with a 0.54 M MobileNetV2, supports zero‑shot deployment on a Jetson Orin Nano UGV, and enables transparent failure analysis under sensor corruption.
By Edward Beng Wai Tan, Siew-Kei Lam
The paper introduces Grace-BEV, a lightweight framework that enhances the robustness of Lift‑Splat‑Shoot (LSS) based bird’s‑eye view (BEV) perception models against sensor corruptions. Grace‑BEV operates in aligned BEV space, combining a LiDAR‑guided image‑BEV expert with a LiDAR‑independent visual expert, and uses a TrustGate Router and FailSafe Fusion Block to mix and recalibrate their outputs. Experiments on nuScenes-R and nuScenes-C show that Grace‑BEV maintains strong performance under diverse corruptions, recovers up to 34.7 mAP when LiDAR is lost, and even improves clean accuracy by up to 1.4 mAP with only 1.2M additional parameters.
By Haifa Zhang, Yijing Wang, Zhiqiang Zuo
End-to-end multimodal driving has progressed rapidly by fusing camera and LiDAR streams. Existing pipelines remain fragile under asymmetric sensor degradation, where either an entire modality or only...
M3GD introduces a multimodal representation that fuses pre‑trained 2D image and 3D LiDAR foundation models for robotic novel view synthesis, avoiding the need for a separate cross‑modal translator. By projecting LiDAR onto the image latent grid and injecting the resulting geometry‑aware packets via a lightweight residual adapter, the method enhances both RGB and depth synthesis on the GrandTour dataset compared to an image‑only baseline. Ablation studies confirm that pixel‑aligned LiDAR content drives the performance gains, and real‑world deployment on a ground robot demonstrates a tunable quality–cost trade‑off.
By Yang Zhou, Jiuhong Xiao, Shizhao Ye, Long Quang, Carlos Nieto-Granda, Giuseppe Loianno
arXiv:2608. 19536v1 Announce Type: cross Abstract: Learning-based global point cloud registration has achieved remarkable progress, yet its reliance on geometric representations makes existing methods sensitive to variations in point density, scan pattern, viewpoint, and sensor characteristics.
By Eunsoo Im, Junghun Suh, Gyeonggwan Lee, Seunghwan Hong
arXiv:2610.27547v1 Announce Type: new
Abstract: Online LiDAR-camera calibration remains a fundamental yet challenging task in the field of robotics. Recently, deep learning-based data-driven approach...
By Huihua Zhu, Shuyang Zhang, Dongqing Chen, Zhuo Yu, Jianhao Jiao, Yilong Zhu, Siyu Teng, Xiaopin Zhong, Zongze Wu