MIT team shows a way to repair AI-generated 3D models before printing
InstructMesh lets users select part of a generated object and correct defects with text instructions or sliders before fabrication.
From plausible shape to usable object
Researchers at MIT, Google and Northeastern University developed InstructMesh to help people repair AI-generated 3D designs before fabrication, MIT News reported on October 1. A user can select a region of a model and describe a local change, or adjust it with sliders. The team’s paper gives examples such as opening or sealing a void and changing thickness. The system works on an intermediate representation of the generated shape, allowing a targeted edit without rebuilding the whole object in conventional modeling software. Researchers fabricated several example items to demonstrate the workflow.
The paper reports two user studies in which novices found and corrected fabrication-related problems, and participants preferred a mix of natural-language instructions and direct controls. Those findings are evidence for the tested tasks, not proof that every AI-generated object will print correctly or withstand real-world use. MIT News says a version of the work is planned for presentation at a November user-interface conference. InstructMesh is a research prototype with a public paper; the reviewed sources do not announce a widely available design service or independent production reliability results.