MIT Develops System to Enhance CAD Model Generation

Researchers at MIT have created a system named GIFT that enhances the conversion of 2D designs into 3D CAD models using vision-language models. This system improves accuracy and reduces computational costs by learning from its own errors and generating new data for training. The development could streamline rapid prototyping and make AI-driven CAD generation more efficient. The research, presented at the International Conference on Machine Learning, was partially funded by the MIT-IBM Computing Research Lab.