Tong Zhao, Zhiyuan Han, Lin Chen, Bowen Xie, Yunxi Qiao, Bohan Li, Weiqing Xiao, Gen Li, Shu Han, Xuyang Dai, Cheng Bi, Hao Zhao, Chaojian Li · 2026 · arXiv (Cornell University)
DOI: 10.48550/arxiv.2610.26585 · cited by 0 works (OpenAlex) · open-access PDF
La génération d'environnements de conduite fixes réutilisables nécessite à la fois des observations futures contrôlables et une représentation explicite qui prend en charge le rendu à partir de différents points de vue, mais les méthodes existantes récupèrent toujours des scènes explicites à partir de vidéos RGB générées grâce à l'optimisation pour chaque séquence. Pour surmonter ces limitations, nous présentons Drive2Gauss, un framework feedforward qui convertit les requêtes RGB générées conjointement, la profondeur métrique et le flux dynamique en scènes Gaussian 4D dynamiques. Un générateur de vidéo multimodal prédit les trois modalités à partir d'images multiview de référence et de signaux de contrôle futurs. En outre, nous introduisons un
Machine translation (Meta M2M100); the English original follows.
Generating reusable driving environments requires both controllable future observations and an explicit representation that supports rendering from different viewpoints, yet existing methods still recover explicit scenes from generated RGB videos through optimization for each sequence. To overcome these limitations, we present Drive2Gauss, a feedforward framework that converts jointly generated RGB, metric depth, and dynamic flow into dynamic 4D Gaussian scenes. A Multimodal Video Generator predicts the three modalities from reference multiview images and future control signals. Furthermore, we introduce a Gaussian Query Decoder that uses depth to initialize sparse 3D queries and flow to associate the queries across time, and integrates appearance information through spatiotemporal feature aggregation and iterative query refinement. The reconstruction stage operates in a single feedforward pass without optimization for each scene. Extensive results on nuScenes demonstrate that our approach attains an FID of 6.38 and an FVD of 12.43 for future RGB generation, a PSNR of 29.91 dB when reconstructing its generated observations, and supports both direct rendering under lateral camera shifts and downstream video fixing. Code and more visualizations are available at https://zttt04. github.io/Drive2Gauss_page/.
Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese
Topics: Computer science, Artificial intelligence, Rendering (computer graphics), Computer vision, Feed forward, RGB color model
GET https://store.agentexchange.work/gw/papers/verify?title=...&year=... (0.003 USD), fetch this record as JSON with /gw/papers/paper?id=2610.26585, or list what it cites with /gw/papers/references?id=.... Pay per call in USDC (x402) or with a 19 USD card key for 1,000 calls. Catalogue · llms.txtSources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange