# NVIDIA's Cosmos 3 generates camera-controlled views from a single image

> Original title: How do you train a world model to follow camera controls?

- Source: NVIDIA AI
- Published: 2026-11-02T19:06:11.000Z
- Bad Influence: https://badinfluence.news/items/nkyvcr5h9ai3w4gogvph4ti11
- Original: https://x.com/NVIDIAAI/status/2117332461785133545

## Summary

NVIDIA shows Cosmos 3 generating new views from one image as the camera follows a dog, rises, or moves sideways. The post links a technical report describing the post-training recipe for camera-controlled world models.

## Post

How do you train a world model to follow camera controls?

From a single image, Cosmos 3 generates views as the camera follows the dog, rises or moves sideways.

Read the technical report for the post-training recipe → https://research.nvidia.com/labs/cosmos-lab/cosmos-simulation/technical-report.pdf

![](https://badinfluence.news/api/img-proxy?u=https%3A%2F%2Fpbs.twimg.com%2Fmedia%2FHWJCoDYaoAI2zG3.jpg&mode=thumb&exp=1793836800&sig=2d0fba09eaf2d0fe)