UFMF (Ultra-Fast Multi-modal Forecaster) is a real‑time 4D occupancy forecasting model that uses a lightweight diagonal state‑space core and a radar‑guided velocity‑aware dynamic prior to predict future motion. The prior splits future displacement into a linear extrapolation and a learned residual, enabling per‑cell flow fields that carry a compact dynamic latent subspace during rollout. With auxiliary world‑frame dynamic supervision and class‑conditioned vertical‑profile supervision, UFMF achieves 122 FPS on an RTX 4090 and outperforms recent occupancy world models on nuScenes while using far less computational resources.
By Zeping Zhang, Zhuoya Zhao, Robert Lagani\`ere
arXiv:2610.09444v1 Announce Type: new
Abstract: Forecasting future 3D occupancy from surround-view cameras is essential for autonomous driving, yet existing approaches rely on dense voxel or bird's-e...
By Hwanhee Jung, SeungHyeon Kim, Inkyu Koo, Qixing Huang, Sang Ho Yoon, Sangpil Kim
arXiv:2610.33356v1 Announce Type: new
Abstract: Event cameras provide microsecond temporal resolution, high dynamic range, and sparse asynchronous measurements, making them compelling for odometry un...
By Matteo Cascone, David Jensen, Alessandro Pirini, Angelo Stefanini, Pietro Ven\`e, Marta Tintor\'e Gazulla, Javier Hidalgo Carri\`o
arXiv:2608.20974v1 Announce Type: cross
Abstract: Video Joint Embedding Predictive Architecture (V-JEPA) learns powerful spatiotemporal representations from video through self-supervised latent featu...
By Xinlin Wang, Yujiao Xiang, Yuheng Zhou, Jingqi Wang, Minqing Huang, Jiajie Huang, Dongxu Wei, Tingguang Zhou, Xiyang Wang, Gong Chen, Zhi Xu, Feiyang Tan, Hangning Zhou, Mu Yang
arXiv:2605.25059v4 Announce Type: replace
Abstract: Crucial for autonomous exploration, online 3D occupancy prediction and mapping incrementally construct dense spatial representations on the fly. Em...
By Ruoyu Wang, Yong Liu, Jiahan Li, Sheng Tao, Yuhang Lin, Yukai Ma
The paper introduces TrajMamba, a lightweight bidirectional state‑space model that predicts short‑term human trajectories from RGB‑D cameras for mobile robots. It aligns training and deployment pipelines by sharing a processing path for pose extraction, depth back‑projection, and smoothing, and encodes recent hip‑center velocities with stacked bidirectional Mamba blocks before decoding future velocities with a GRU head. On the ETH/UCY benchmark, TrajMamba achieves an average ADE of 0.29 m and FDE of 0.51 m, runs in real time on modest hardware, and has been deployed on a TIAGo robot equipped with an NVIDIA Jetson Orin.
By Basheer Al-Tawil, Magnus Jung, Ayoub Al-Hamadi
arXiv:2607. 29235v1 Announce Type: cross Abstract: Although world-action models (WAMs) enhance long-horizon robot control by predicting visual evolution before acting, long-horizon reliability demands repeated re-grounding in real observations--not recursive rollout.
By Peize Li, Ruimeng Zhang, Ru Zhang, Cong Huang, Kai Chen, Shanghang Zhang
arXiv:2606. 12987v1 Announce Type: cross Abstract: Action-conditioned world models let an autonomous vehicle predict future camera scenes from its own planned controls, enabling planning and simulation without real-world rollouts, but at compact, trainable scale the futures are ambiguous and the field's standard distortion metrics actively mislead: they reward a blurry regression mean over a realistic prediction.
By Ruslan Sharifullin, Benjamin Jiang, Kai Xi Chew
RoadOcc is a new method for roadside occupancy prediction that learns to route information among three memory sources: Persist (fixed-coordinate history), Transport (velocity-addressed history), and Refresh (current evidence). It employs dynamic-aware cross‑attention, multi‑scale voxel velocity estimation, and velocity‑guided dynamic sparse fusion to combine these sources efficiently. On the InfraOcc dataset, RoadOcc achieves 65.29 mIoU and 32.37 dynamic mIoU, outperforming the previous STCOcc baseline by significant margins.
By Xiaokai Bai, Lei Yang, Songkai Wang, Lianqing Zheng, Si-Yuan Cao, Hui-liang Shen
Vision-Language-Action (VLA) models generalize across tasks, but their autoregressive decoders need sequential forward passes in proportion to the action horizon, each bounded below on embedded aerial...
PhysWAM is a unified world-action model for autonomous driving that jointly denoises multiview video, metric depth, and ego motion using a flow‑matching transformer. It introduces Coupled Point Projection (CPP), a geometric constraint that aligns generated depth points with LiDAR data after applying the predicted SE(3) ego motion, thereby enforcing physical consistency. At inference, trajectory selection uses a simple label‑free consensus rule, and the model demonstrates strong planning performance, zero‑shot transfer to unseen environments, and accurate, temporally coherent depth and video predictions.
By Dhruv Parikh, Fengcheng Yu, Quankai Gao, Jiawei Yang, Junjie Ye, Maulik Bhatt, Thang Vu, Charles Ochoa, Rowan McAllister, Igor Vasiljevic, Rajgopal Kannan, Viktor Prasanna, Vitor Guizilini, Yue Wang
Kairos extends a hierarchical 3D scene graph to a 4D scene graph, storing for each voxel a directional mixture and a presence rate. Spectral predictors forecast both the probability of people being present and the full directional distribution of their motion at any future query time. The model supports conditional queries via pairwise flow dependence and provides calibrated credible intervals that tighten as observations accumulate.
By Iacopo Catalano, Julio A. Placed, Javier Civera, Jorge Pe\~na Queralta