用多层特征融合提升工业质检模型持续学习能力
Multi-Level Feature Fusion for Continual Learning in Visual Quality Inspection
- 从预训练网络不同层级提取特征进行融合
- 参数量更少，性能接近端到端训练
- 有效缓解灾难性遗忘，适合新产品/缺陷场景
深度神经网络在制造领域的视觉质量检测中展现出巨大潜力。然而，在再制造等动态场景中，被检产品和缺陷模式常发生变化，导致部署模型需频繁适应新条件，形成持续学习问题。为实现快速适应，训练过程必须计算高效，同时避免灾难性遗忘。本文提出多层级特征融合（MLFF）方法，利用预训练网络不同深度的表示，同时提升效率与鲁棒性。实验表明，该方法在多种质量检测任务上性能接近端到端训练，但可训练参数显著减少；同时有效降低灾难性遗忘，增强对新型产品或缺陷的泛化能力。
原文摘要 · Abstract (English)
Deep neural networks show great potential for automating various visual quality inspection tasks in manufacturing. However, their applicability is limited in more volatile scenarios, such as remanufacturing, where the inspected products and defect patterns often change. In such settings, deployed models require frequent adaptation to novel conditions, effectively posing a continual learning problem. To enable quick adaptation, the necessary training processes must be computationally efficient while still avoiding effects like catastrophic forgetting. This work presents a multi-level feature fusion (MLFF) approach that aims to improve both aspects simultaneously by utilizing representations from different depths of a pretrained network. We show that our approach is able to match the performance of end-to-end training for different quality inspection problems while using significantly less trainable parameters. Furthermore, it reduces catastrophic forgetting and improves generalization robustness to new product types or defects.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。