# GeoStream: Toward Precise Camera Controlled Streaming Video Generation

arXiv: https://arxiv.org/abs/2606.15162 · HTML: https://mlsanity.com/paper/2606.15162/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- Placed in Video-generation world models and simulators, one of 11 directions in World models.
- Builds on 35 of its 84 references that are in the map.
- First posted on arXiv 13 Jun 2026.

## Placed in

- World models / [Video-generation world models and simulators](https://mlsanity.com/wm/video-generation-world-models-and-simulators/)

## Builds on (most cited first, 30 of 35)

- [Wan: Open and Advanced Large-Scale Video Generative Models](https://mlsanity.com/paper/2503.20314/) (2025)
- [Genie: Generative Interactive Environments](https://mlsanity.com/paper/2402.15391/) (2024)
- [MotionCtrl: A Unified and Flexible Motion Controller for Video Generation](https://mlsanity.com/paper/2312.03641/) (2023)
- [World Action Models are Zero-shot Policies](https://mlsanity.com/paper/2602.15922/) (2026)
- [ViewCrafter: Taming Video Diffusion Models for High-fidelity Novel View Synthesis](https://mlsanity.com/paper/2409.02048/) (2024)
- [π3: Permutation-Equivariant Visual Geometry Learning](https://mlsanity.com/paper/2507.13347/) (2025)
- [MAGI-1: Autoregressive Video Generation at Scale](https://mlsanity.com/paper/2505.13211/) (2025)
- [Video Depth Anything: Consistent Depth Estimation for Super-Long Videos](https://mlsanity.com/paper/2501.12375/) (2025)
- [Self-Forcing++: Towards Minute-Scale High-Quality Video Generation](https://mlsanity.com/paper/2510.02283/) (2025)
- [Diffusion as Shader: 3D-aware Video Diffusion for Versatile Video Generation Control](https://mlsanity.com/paper/2501.03847/) (2025)
- [Causal Forcing: Autoregressive Diffusion Distillation Done Right for High-Quality Real-Time Interactive Video Generation](https://mlsanity.com/paper/2602.02214/) (2026)
- [UniDepthV2: Universal Monocular Metric Depth Estimation Made Simpler](https://mlsanity.com/paper/2502.20110/) (2025)
- [Advancing Open-source World Models](https://mlsanity.com/paper/2601.20540/) (2026)
- [Video World Models with Long-term Spatial Memory](https://mlsanity.com/paper/2506.05284/) (2025)
- [VGGT-Long: Chunk it, Loop it, Align it -- Pushing VGGT's Limits on Kilometer-scale Long RGB Sequences](https://mlsanity.com/paper/2507.16443/) (2025)
- [Geo4D: Leveraging Video Generators for Geometric 4D Scene Reconstruction](https://mlsanity.com/paper/2504.07961/) (2025)
- [World-consistent Video Diffusion with Explicit 3D Modeling](https://mlsanity.com/paper/2412.01821/) (2024)
- [InfiniteVGGT: Visual Geometry Grounded Transformer for Endless Streams](https://mlsanity.com/paper/2601.02281/) (2026)
- [Memory Forcing: Spatio-Temporal Memory for Consistent Scene Generation on Minecraft](https://mlsanity.com/paper/2510.03198/) (2025)
- [FlexWorld: Progressively Expanding 3D Scenes for Flexiable-View Synthesis](https://mlsanity.com/paper/2503.13265/) (2025)
- [World-VLA-Loop: Closed-Loop Learning of Video World Model and VLA Policy](https://mlsanity.com/paper/2602.06508/) (2026)
- [EX-4D: EXtreme Viewpoint 4D Video Synthesis via Depth Watertight Mesh](https://mlsanity.com/paper/2506.05554/) (2025)
- [FlashWorld: High-quality 3D Scene Generation within Seconds](https://mlsanity.com/paper/2510.13678/) (2025)
- [LoL: Longer than Longer, Scaling Video Generation to Hour](https://mlsanity.com/paper/2601.16914/) (2026)
- [The Role of World Models in Shaping Autonomous Driving: A Comprehensive Survey](https://mlsanity.com/paper/2502.10498/) (2025)
- [VerseCrafter: Dynamic Realistic Video World Model with 4D Geometric Control](https://mlsanity.com/paper/2601.05138/) (2026)
- [Scalable Policy Evaluation with Video World Models](https://mlsanity.com/paper/2511.11520/) (2025)
- [Taming Video Models for 3D and 4D Generation via Zero-Shot Camera Control](https://mlsanity.com/paper/2509.15130/) (2025)
- [WorldWeaver: Generating Long-Horizon Video Worlds via Rich Perception](https://mlsanity.com/paper/2508.15720/) (2025)
- [Latent-Reframe: Enabling Camera Control for Video Diffusion Model without Training](https://mlsanity.com/paper/2412.06029/) (2024)

## Built on by (most cited first, 4 of 4)

- [MemoBench: Benchmarking World Modeling in Dynamically Changing Environments](https://mlsanity.com/paper/2606.27537/) (2026)
- [From RGB Generation to Dense Field Readout: Pixel-Space Dense Prediction with Text-to-Image Models](https://mlsanity.com/paper/2607.06553/) (2026)
- [FlashRender: Few-Step Generative Rendering via Camera-Controlled Video MeanFlow](https://mlsanity.com/paper/2609.03563/) (2026)
- [Honeycomb: Constant-Size Scene Memory Representation for Video World Models](https://mlsanity.com/paper/2609.37690/) (2026)

## Abstract (arXiv)

Accurate interactive camera control is essential for video-based world models, but most existing approaches learn camera motion implicitly, leading to inaccurate control under out-of-distribution trajectories. Explicit geometric conditioning improves controllability, but existing methods are non-autoregressive and rely on a static 3D cache built from an initial frame, which becomes ineffective once the viewpoint moves beyond the original frustum. We propose GeoStream, a framework that enables precise metric-scale camera control in autoregressive streaming video generation. Our method maintains a self-refreshing 3D cache that is periodically updated online from the model's own outputs: we estimate depth from the most recently generated frame, unproject to 3D, and reproject into the target view to produce point reprojections as geometric conditioning for subsequent synthesis. By the same principle, the conditioning seen during training is also rendered from the student's own generated frames, yielding a fully on-policy distillation that naturally aligns the train and inference conditioning distributions. Unlike prior work that uses off-policy condition noising, our approach trains the model against the exact error distribution it encounters at inference, mitigating both standard autoregressive drift and the second-order geometric feedback loop that arises when the cache itself is derived from generated outputs. Quantitative and qualitative results show that our approach substantially improves camera controllability.