Generate Cinematic Video Environments Using Street View Data

Google has introduced a significant update to Veo, its flagship video generation model, by integrating data from Street View and Project Genie. This update allows creators to simulate real-world locations with high spatial accuracy, moving beyond generic AI-generated landscapes into identifiable, geographically grounded environments. For filmmakers and virtual production teams, this provides a bridge between location scouting and digital asset creation.

What's new

The primary update is the ability to use Street View’s vast library of panoramic imagery as a foundation for video synthesis. Instead of relying solely on text prompts to describe a scene, the model can now reference the physical layout, architecture, and lighting of actual coordinates. This results in videos that maintain structural integrity and perspective as the virtual camera moves through the space.

Key technical improvements include:

• Improved spatial consistency when simulating camera movement through 3D environments.
• The ability to generate multi-angle views of a single location while maintaining consistent textures and geometry.
• Integration with Project Genie, which focuses on interactive environment simulation, allowing for more control over how a character or camera interacts with the generated world.

How it fits your workflow

This update changes how directors and visual effects artists approach pre-visualization and background plate generation. Traditionally, if a production needed a specific street in London or a highway in California but couldn't film on-site, they would rely on expensive stock footage or manual 3D environment builds. Google Veo now allows for the generation of these plates using real-world data, which can be customized for specific lighting conditions or time of day.

For location scouts, this serves as a digital sandbox. You can take a real-world coordinate and test how different camera paths might look before ever sending a crew to the physical site. While tools like Sora or Runway Gen-3 Alpha offer high-fidelity visuals, the integration of Street View gives Veo a unique advantage in geographic accuracy. It functions less like a creative paintbrush and more like a programmable drone that can fly through a digital twin of the planet.

Editors and VFX artists can use these generated clips as high-quality placeholders or even final background elements for green-screen compositions. The consistency of the 3D space makes it easier to track movement and composite actors into the scene compared to standard generative video which often suffers from warping or "hallucinated" architecture.

What it costs / how to try it

Google is expanding access to these new capabilities for Google One AI Premium subscribers globally. Creators can test the features through VideoFX, Google’s experimental platform for creative tools. Availability may vary by region as the rollout continues across different markets.