Do generative video models understand physical principles?
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866929733630951424 |
|---|---|
| author | Motamed, Saman Culp, Laura Swersky, Kevin Jaini, Priyank Geirhos, Robert |
| author_facet | Motamed, Saman Culp, Laura Swersky, Kevin Jaini, Priyank Geirhos, Robert |
| contents | AI video generation is undergoing a revolution, with quality and realism advancing rapidly. These advances have led to a passionate scientific debate: Do video models learn "world models" that discover laws of physics -- or, alternatively, are they merely sophisticated pixel predictors that achieve visual realism without understanding the physical principles of reality? We address this question by developing Physics-IQ, a comprehensive benchmark dataset that can only be solved by acquiring a deep understanding of various physical principles, like fluid dynamics, optics, solid mechanics, magnetism and thermodynamics. We find that across a range of current models (Sora, Runway, Pika, Lumiere, Stable Video Diffusion, and VideoPoet), physical understanding is severely limited, and unrelated to visual realism. At the same time, some test cases can already be successfully solved. This indicates that acquiring certain physical principles from observation alone may be possible, but significant challenges remain. While we expect rapid advances ahead, our work demonstrates that visual realism does not imply physical understanding. Our project page is at https://physics-iq.github.io; code at https://github.com/google-deepmind/physics-IQ-benchmark. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2501_09038 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Do generative video models understand physical principles? Motamed, Saman Culp, Laura Swersky, Kevin Jaini, Priyank Geirhos, Robert Computer Vision and Pattern Recognition Artificial Intelligence Graphics Machine Learning AI video generation is undergoing a revolution, with quality and realism advancing rapidly. These advances have led to a passionate scientific debate: Do video models learn "world models" that discover laws of physics -- or, alternatively, are they merely sophisticated pixel predictors that achieve visual realism without understanding the physical principles of reality? We address this question by developing Physics-IQ, a comprehensive benchmark dataset that can only be solved by acquiring a deep understanding of various physical principles, like fluid dynamics, optics, solid mechanics, magnetism and thermodynamics. We find that across a range of current models (Sora, Runway, Pika, Lumiere, Stable Video Diffusion, and VideoPoet), physical understanding is severely limited, and unrelated to visual realism. At the same time, some test cases can already be successfully solved. This indicates that acquiring certain physical principles from observation alone may be possible, but significant challenges remain. While we expect rapid advances ahead, our work demonstrates that visual realism does not imply physical understanding. Our project page is at https://physics-iq.github.io; code at https://github.com/google-deepmind/physics-IQ-benchmark. |
| title | Do generative video models understand physical principles? |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence Graphics Machine Learning |
| url | https://arxiv.org/abs/2501.09038 |