Process uncertainty has a significant impact on industrial image processing. Existing deep learning methods were established on high-quality datasets without considering the uncertainty. This paper proposes lightweight spatial-channel feature disentanglement modeling with confidence evaluation for uncertain industrial images. First, spatial-channel feature disentanglement modeling inspired by tensor decomposition aims to balance the computational efficiency and feature expression capabilities. Second, collaborative learning with confidence evaluation is designed to cope with uncertain samples. Then, representative features are fine-tuned on high-confidence datasets for optimal performance. Complexity analysis and experiments verified the effectiveness, computational efficiency, and robustness of the proposed model.

中文翻译：

---

面向不确定工业图像的轻量级空间通道特征解缠建模与置信度评估


过程不确定性对工业图像处理有重大影响。现有的深度学习方法是在高质量数据集上建立的，没有考虑不确定性。本文提出了一种轻量级空间通道特征解纠缠建模，并针对不确定的工业图像进行置信度评估。首先，受张量分解启发的时空通道特征解纠建模旨在平衡计算效率和特征表达能力。其次，具有置信度评估的协作学习旨在应对不确定的样本。然后，在高置信度数据集上微调代表性特征以获得最佳性能。复杂性分析和实验验证了所提模型的有效性、计算效率和鲁棒性。