Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

SPFFNet: Strip Perception and Feature Fusion Spatial Pyramid Pooling for Fabric Defect Detection

Peizhe Zhao, Shunbo Jia

TL;DR

This paper tackles fabric defect detection under complex backgrounds by introducing SPFFNet, a YOLOv11-based detector augmented with a Strip Perception Module (SPM), SE-SPPF for fused spatial-channel features, and FECIoU for robust localization across varied defect scales. The proposed components collectively enhance stripe defect perception, multi-scale feature fusion, and localization accuracy while maintaining inference speed. Empirical results on Tianchi and a custom dataset show state-of-the-art mAP improvements (65.8% and 90.6%, respectively), with ablative evidence supporting each module's contribution. The work advances industrial fabric inspection by delivering a high-accuracy, real-time capable framework, with future directions including cross-domain and spectral-feature integration to broaden generalization.

Abstract

Defect detection in fabrics is critical for quality control, yet existing methods often struggle with complex backgrounds and shape-specific defects. In this paper, we propose an improved fabric defect detection model based on YOLOv11. To enhance the detection of strip defects, we introduce a Strip Perception Module (SPM) that improves feature capture through multi-scale convolution. We further enhance the spatial pyramid pooling fast (SPPF) by integrating a squeeze-and-excitation mechanism, resulting in the SE-SPPF module, which better integrates spatial and channel information for more effective defect feature extraction. Additionally, we propose a novel focal enhanced complete intersection over union (FECIoU) metric with adaptive weights, addressing scale differences and class imbalance by adjusting the weights of hard-to-detect instances through focal loss. Experimental results demonstrate that our model achieves a 0.8-8.1% improvement in mean average precision (mAP) on the Tianchi dataset and a 1.6-13.2% improvement on our custom dataset, outperforming other state-of-the-art methods.

SPFFNet: Strip Perception and Feature Fusion Spatial Pyramid Pooling for Fabric Defect Detection

TL;DR

This paper tackles fabric defect detection under complex backgrounds by introducing SPFFNet, a YOLOv11-based detector augmented with a Strip Perception Module (SPM), SE-SPPF for fused spatial-channel features, and FECIoU for robust localization across varied defect scales. The proposed components collectively enhance stripe defect perception, multi-scale feature fusion, and localization accuracy while maintaining inference speed. Empirical results on Tianchi and a custom dataset show state-of-the-art mAP improvements (65.8% and 90.6%, respectively), with ablative evidence supporting each module's contribution. The work advances industrial fabric inspection by delivering a high-accuracy, real-time capable framework, with future directions including cross-domain and spectral-feature integration to broaden generalization.

Abstract

Defect detection in fabrics is critical for quality control, yet existing methods often struggle with complex backgrounds and shape-specific defects. In this paper, we propose an improved fabric defect detection model based on YOLOv11. To enhance the detection of strip defects, we introduce a Strip Perception Module (SPM) that improves feature capture through multi-scale convolution. We further enhance the spatial pyramid pooling fast (SPPF) by integrating a squeeze-and-excitation mechanism, resulting in the SE-SPPF module, which better integrates spatial and channel information for more effective defect feature extraction. Additionally, we propose a novel focal enhanced complete intersection over union (FECIoU) metric with adaptive weights, addressing scale differences and class imbalance by adjusting the weights of hard-to-detect instances through focal loss. Experimental results demonstrate that our model achieves a 0.8-8.1% improvement in mean average precision (mAP) on the Tianchi dataset and a 1.6-13.2% improvement on our custom dataset, outperforming other state-of-the-art methods.

Paper Structure

This paper contains 21 sections, 5 equations, 4 figures, 3 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Object Detection
  4. Fabric Defect Detection Algorithms
  5. Attention Mechanism
  6. Proposed Method
  7. Overview
  8. Strip Perception Module
  9. Squeeze and Excitation Spatial Pyramid Pooling Fast
  10. Focal Enhanced Complete Intersection over Union
  11. Experiment and Analysis
  12. Datasets
  13. Tianchi fabric dataset
  14. Produced dataset
  15. Implementation Details
  16. ...and 6 more sections

Figures (4)

  • Figure 1: Network structure of the proposed method
  • Figure 2: (a) illustrates the overall architecture of the Strip Perception Module, while (b) presentsdetailed fusion operations within the Dense Connections component. The Dense Connectionsfacilitate comprehensive integration of diverse strip-wise features.
  • Figure 3: (a) illustrates the overall architecture of the Squeeze-and-Excitation Spatial Pyramid Pooling Fast (SE-SPPF), while (b) provides a detailed breakdown of the SENetV2 module's processing pipeline. The SENetV2 module effectively harmonizes multi-scale features to extract the most discriminative characteristics for defect detection.
  • Figure 4: Comparison visualized by heat maps