Dimensions overall score 8.0

GitHub Code Pulse

No public code linked for this paper yet.

• CHROMM jointly estimates cameras, scene point clouds, and human meshes from multi-person multi-view videos without relying on external modules or preprocessing.method 95%
• CHROMM integrates priors from Pi3X and Multi-HMR into a single trainable neural network architecture.method 90%
• CHROMM introduces a scale adjustment module to solve the scale discrepancy between humans and the scene.technical 90%
• CHROMM's geometry-based multi-person association method is more robust than appearance-based approaches.technical 85%
• CHROMM runs over 8x faster than prior optimization-based multi-view approaches.result 90%
• CHROMM achieves competitive performance in global human motion and multi-view pose estimation on datasets including EMDB, RICH, EgoHumans, and EgoExo4D.result 85%
• Most existing approaches for reconstructing humans and environments focus on monocular inputs.limitation 80%
• Extending monocular approaches to multi-view settings requires additional overhead modules or preprocessed data.limitation 75%

Competitive landscape

Competitor map is still being generated for this paper. Enable generation or check back soon.

Startup potential card