Synthetic data framework automates gravure printing defect detection

Researchers have developed a new framework for generating synthetic data to automate quality control in gravure printing. This approach addresses the scarcity of real-world defect images, which hinders the training of deep learning models like YOLO and Vision Transformers. The framework creates high-fidelity images of printing defects along with corresponding annotations, enabling the training of robust models without extensive manual data collection. When tested with the RFDETR model, synthetic data yielded an 80.9% mAP on real industrial samples, offering a cost-effective and rapid solution for defect inspection. AI

IMPACT This framework could significantly reduce the cost and time required for quality control in manufacturing by enabling more efficient training of AI defect detection models.

RANK_REASON The cluster contains an academic paper detailing a new synthetic data generation framework for a specific industrial application. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• arXiv
• Hugging Face
• RFDETR
• Vision Transformers
• YOLO

• paper
• product

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →