Drive3R: Block-wise Streaming Geometric Foundation Models for Autonomous Driving

arXiv:2610.15112 · cs.CV · Submitted 2026-10-10 · Read on arXiv

cs.CV

Submitted: 2026-10-10

Updated: 2026-10-10

Comments: 11 pages, 6 figures, 4 tables. Project page: https://silence-ai423.github.io/Drive3R/

Project page: https://silence-ai423.github.io/Drive3R

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• Foundation Model — 1× in this paper · explained in 3 episodes
• Foundation Models — 1× in this paper · explained in 7 episodes

Sources

• TTT3R: 3D Reconstruction as Test-Time Training
• LongStream: Long-Sequence Streaming Autoregressive Visual Geometry
• VGGT-Long: Chunk it, Loop it, Align it -- Pushing VGGT's Limits on Kilometer-scale Long RGB Sequences
• LRM: Large Reconstruction Model for Single Image to 3D
• Depth Anything 3: Recovering the Visual Space from Any Views
• FastVGGT: Training-Free Acceleration of Visual Geometry Transformer
• UniSplat: Unified Spatio-Temporal Fusion via 3D Latent Scaffolds for Dynamic Driving Scene Reconstruction
• $\pi^3$: Permutation-Equivariant Visual Geometry Learning
• ReconDrive: Fast Feed-Forward 4D Gaussian Splatting for Autonomous Driving Scene Reconstruction
• InfiniteVGGT: Visual Geometry Grounded Transformer for Endless Streams
• Efficient and Explainable End-to-End Autonomous Driving via Masked Vision-Language-Action Diffusion
• Review of Feed-forward 3D Reconstruction: From DUSt3R to VGGT
• OneDrive: Unified Multi-Paradigm Driving with Vision-Language-Action Models
• AutoVLA: A Vision-Language-Action Model for End-to-End Autonomous Driving with Adaptive Reasoning and Reinforcement Fine-Tuning
• Streaming 4D Visual Geometry Transformer
• DVGT: Driving Visual Geometry Transformer
• DVGT-2: Vision-Geometry-Action Model for Autonomous Driving at Scale

Related papers

• Loss Knows Best: Detecting Annotation Errors in Videos via Loss Trajectories
• AnchorWeave: World-Consistent Video Generation with Retrieved Local Spatial Memories
• Benchmarking the Robustness of Foundation Models for Mammography under Domain Shift
• MambaX-Net: Dual-Input Mamba-Enhanced Cross-Attention Network for Longitudinal MRI Segmentation
• TeleOCR: Navigating Document Parsing Across Digital and Camera-Captured Documents
• A Survey on Efficient Vision-Language-Action Models