Neural reconstruction has advanced significantly in the past year, and dynamic models are becoming increasingly common. However, these models are limited to handling in-domain objects that closely follow their original trajectories. 

This blog post presents a hybrid approach that integrates the advantages of neural reconstruction with physics-based rendering. 

First, we remove dynamic objects from the scene and reconstruct the static environment using a neural reconstruction model. Then, we populate the reconstructed environment with dynamic objects in aiSim. This approach effectively mitigates the drawbacks of both methods—such as domain gap in traditional simulation and out-of-domain object rendering in neural reconstruction.

See details at our CVPR2025 subpage