ACADEMICNeurIPS Spotlight · 2024

Text2CAD

Text2CAD generates parametric CAD models from natural language descriptions using a transformer trained on the DeepCAD dataset. The model maps text prompts at four abstraction levels (basic shape, part-level, feature-level, dimension-level) to sketch-extrude sequences. NeurIPS 2024 Spotlight.

Input
Text
Output
CAD sequences
Venue
NeurIPS Spotlight
Year
2024
Links
GitHub →Project page →
CAD Arena Status
Scheduled for evaluation on the full 200-prompt benchmark. Results will appear on the leaderboard at launch.