Visual Quality Score Assessment of Large White Goods in Remanufacture with Multi-View Deformable-DETR

2026-06-12

AI Summary

This work addresses the challenges of manual visual inspection in large white goods remanufacturing, where low efficiency, high annotation costs, and difficulty in detecting subtle defects in multi-view high-resolution images hinder quality control. To overcome these limitations, the authors propose an automated quality scoring framework based on multi-view Deformable DETR that fuses redundant information across views to extract fine-grained features. By integrating self-supervised pretraining with expert-score-guided fine-tuning, the method achieves high-precision assessment while significantly reducing reliance on labeled data. A novel linear projection mechanism over frozen feature maps is introduced to enable interpretable localization of defective regions. Experimental validation on an industrial dataset demonstrates the effectiveness of the approach, advancing scalable and transparent quality inspection in remanufacturing.

Abstract

Remanufacturing large white goods is essential for a circular economy, yet visual quality assessment remains a manual bottleneck for training and pricing. Conventional detection methods require extensive annotation and struggle with small defects in high-resolution multi-view data. We present a multi-view framework based on Deformable-DETR for automated quality scoring that aggregates information across redundant views to extract fine-grained features. To enhance robustness with limited labels, we employ self-supervised pretraining followed by supervised fine-tuning on expert-annotated scores. Additionally, a linear projection over frozen feature maps identifies regions of interest to explain model decisions. Evaluated on an industrial multi-view dataset, our approach delivers precise quality assessments while reducing reliance on manual annotation and per-part customization, enabling scalable and transparent inspection for remanufacturing lines.

Problem

• visual quality assessment
• remanufacturing
• multi-view
• small defect detection
• manual bottleneck

Innovation

• Multi-View Deformable-DETR
• Self-Supervised Pretraining
• Visual Quality Scoring
• Explainable AI
• Remanufacturing Inspection