CINFormer: Transformer network with multi-stage CNN feature injection for surface defect segmentation

TL;DR

CINFormer is a novel transformer-CNN hybrid network designed for surface defect segmentation, effectively combining detailed feature extraction with noise suppression to achieve state-of-the-art results in industrial defect detection.

Contribution

The paper introduces CINFormer, a UNet-like architecture with multi-stage CNN feature injection and a Top-K self-attention module for improved defect segmentation.

Findings

Achieves state-of-the-art performance on DAGM 2007, Magnetic tile, and NEU datasets.

Effectively combines CNN and transformer strengths for detailed and noise-robust defect detection.

Outperforms existing methods in surface defect segmentation accuracy.

Abstract

Surface defect inspection is of great importance for industrial manufacture and production. Though defect inspection methods based on deep learning have made significant progress, there are still some challenges for these methods, such as indistinguishable weak defects and defect-like interference in the background. To address these issues, we propose a transformer network with multi-stage CNN (Convolutional Neural Network) feature injection for surface defect segmentation, which is a UNet-like structure named CINFormer. CINFormer presents a simple yet effective feature integration mechanism that injects the multi-level CNN features of the input image into different stages of the transformer network in the encoder. This can maintain the merit of CNN capturing detailed features and that of transformer depressing noises in the background, which facilitates accurate defect detection. In addition, CINFormer presents a Top-K self-attention module to focus on tokens with more important information about the defects, so as to further reduce the impact of the redundant background. Extensive experiments conducted on the surface defect datasets DAGM 2007, Magnetic tile, and NEU show that the proposed CINFormer achieves state-of-the-art performance in defect detection.