robotics
A Collision Reveals the Cavity You Cannot See
Instead of optimizing invisible density directly, Structure from Collision searches explicit hypotheses about an object's hidden interior.

Summary
Instead of optimizing invisible density directly, Structure from Collision searches explicit hypotheses about an object's hidden interior.
The method watches multiview deformation during a collision, evaluates candidate cavity structures with forward-only material-point simulation and refines the winning hypothesis with differentiable simulation. On the paper's benchmark it improved hidden-structure reconstruction and future prediction over a density-based baseline. The result is a simulated inference method with documented stress tests, not a general nondestructive-inspection instrument.
Why it matters
Instead of optimizing invisible density directly, Structure from Collision searches explicit hypotheses about an object's hidden interior.
Limits and context
- The method watches multiview deformation during a collision, evaluates candidate cavity structures with forward-only material-point simulation and refines the winning hypothesis with differentiable simulation.
- The result is a simulated inference method with documented stress tests, not a general nondestructive-inspection instrument.
Key claims
Instead of optimizing invisible density directly, Structure from Collision searches explicit hypotheses about an object's hidden interior.
Qualification: The method watches multiview deformation during a collision, evaluates candidate cavity structures with forward-only material-point simulation and refines the winning hypothesis with differentiable simulation.
Evidence: source-2026-11-07-013
Sources
- arXiv preprint 2611.07900, original v1arXiv · primary research
Corrections
: Publication note: This November 7 edition was assembled before its scheduled 9:00 AM America/Detroit reveal but is being published late after recovery from a local build interruption. Its original reporting, image-generation times and scheduled publishAt are preserved.