DefectFill: Realistic Defect Generation with Inpainting Diffusion Model for Visual Inspection
This addresses data scarcity for visual inspection in industrial applications, though it is incremental as it builds on existing diffusion models.
The paper tackles the problem of scarce defect data for visual inspection by proposing DefectFill, a method that generates realistic defect images using a fine-tuned inpainting diffusion model with custom loss functions and a low-fidelity selection technique, resulting in state-of-the-art performance on the MVTec AD dataset.
Developing effective visual inspection models remains challenging due to the scarcity of defect data. While image generation models have been used to synthesize defect images, producing highly realistic defects remains difficult. We propose DefectFill, a novel method for realistic defect generation that requires only a few reference defect images. It leverages a fine-tuned inpainting diffusion model, optimized with our custom loss functions incorporating defect, object, and attention terms. It enables precise capture of detailed, localized defect features and their seamless integration into defect-free objects. Additionally, our Low-Fidelity Selection method further enhances the defect sample quality. Experiments show that DefectFill generates high-quality defect images, enabling visual inspection models to achieve state-of-the-art performance on the MVTec AD dataset.