Flash framework speeds up synthetic anomaly creation for industry defects
FLASH: A "Generate Once, Synthesize Many" Framework for Synthetic Anomaly Generation in Industrial Anomaly Detection
Computer Vision and Pattern Recognition
Summary
Detecting industrial defects often needs lots of examples, but real defect images are rare. The authors propose FLASH, which creates a small set of defect images once and then uses these pieces repeatedly to make many new defect images quickly. This approach mixes detected defect patches onto new normal images with varied size and placement, making synthetic defects that help train detection systems. FLASH performs nearly as well as real defects in tests and runs much faster than previous methods that generate every defect from scratch.
What this means in practice
- •For industrial quality assurance teams: Create large, diverse datasets of synthetic defect images to train and calibrate industrial anomaly detection systems when real defect data is limited.
- •For manufacturing ai developers: Generate varied synthetic anomalies efficiently to improve machine learning models for automated defect detection across different products and settings.
Authors
Abhay Kumar Das, Rajesh Gangireddy, Ashwin Vaidya, Samet Akcay
Abstract
Synthetic anomaly generation helps expand industrial anomaly datasets when real defects are scarce or unavailable. Existing approaches lie at two extremes: procedural approaches are fast but struggle to represent complex anomalies, while generative approaches produce diverse defects but require costly per-sample generation. We present FLASH, a framework that decouples defect generation from anomaly synthesis under a ``generate once, synthesize many'' paradigm. Given only normal images, FLASH uses Vision-Language Model (VLM) guidance and an image-generation model to produce a small set of defect images, from which it extracts, validates, and banks reusable defect patches. For synthesis of anomalous images, Object Boundary Suppression (OBS) first identifies the probable foreground object-aware region of the host image, while Multi-Resolution Spectral Pyramid (MRSP) noise generates diverse, size-controllable masks that determine the defect location and spatial extent. It then composes a large and diverse synthetic anomalous image set by localizing the defect region, sampling size-controllable placement masks and seamlessly blending retrieved defects onto new defect-free images without further need for image generation. Experiments on the MVTec AD 2 dataset show that FLASH-generated anomalies nearly close the calibration gap on real defects, reaching 78.1% image-level F1 against an 83.6% real-anomaly upper bound and providing the most consistent calibration transfer across detectors among procedural and generative alternatives. Moreover, FLASH synthesizes anomalies more than 11.95x faster than per-sample generative approaches.