Bridging the Sim-to-Real Gap in Semiconductor Visual Program Synthesis via Input Binarization
Researchers have developed a novel visual program synthesis framework to address the sim-to-real gap in semiconductor inspection. This approach uses a Vision-Language Model (VLM) to translate inspection images into editable code, allowing for precise geometric control and data generation. By employing an input binarization strategy, the model can effectively strip away noise and texture from real-world SEM images, enabling it to focus on crucial geometric structures and improving accuracy. AI
IMPACT This research could lead to more accurate and efficient semiconductor inspection systems by improving the training of AI models on real-world data.