Summary of Ddnet: Deformable Convolution and Dense Fpn For Surface Defect Detection in Recycled Books, by Jun Yu et al.
DDNet: Deformable Convolution and Dense FPN for Surface Defect Detection in Recycled Books
by Jun Yu, WenJian Wang
First submitted to arxiv on: 8 Sep 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Artificial Intelligence (cs.AI)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The proposed DDNet model is designed to enhance defect localization and classification in recycled and recirculated books. The model introduces a deformable convolution operator (DC) and densely connected FPN module (DFPN) to improve boundary delineation, prediction accuracy, and feature fusion. The DC module dynamically adjusts the convolution grid to capture subtle shape variations, while DFPN leverages dense skip connections to generate multi-resolution feature maps. A comprehensive dataset is presented for surface defect detection, featuring a diverse range of defect types, shapes, and sizes. DDNet achieves precise localization and classification, outperforming the baseline model with a mAP value of 46.7% on the proprietary dataset. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper proposes a new way to detect defects in old books. The method uses special computer vision tools to find tiny cracks or damage on book covers. This is important because old books are valuable and people need to be able to check them for damage before buying or selling them. The team created a special dataset with many different types of defects, which helps the model learn how to detect them accurately. The new method works better than previous ones, making it useful for book collectors and sellers. |
Keywords
» Artificial intelligence » Classification
