Learning to utilize image second-order derivative information for crisp edge detection

2024-06-09

arXiv.org

AI Summary

To address challenges in edge detection—including thick and noisy edges, unknown edge regularities, and imbalanced pixel distributions—this paper proposes LUS-Net. First, it introduces a geometric prior by explicitly modeling the image’s second-order derivatives to guide precise edge localization. Second, it designs the Scale-Dependent Multi-Context Module (SDMCM) to jointly fuse multi-scale contextual information and second-order geometric features. Third, it adopts a Hybrid Focal Loss (HFL) to mitigate foreground–background class imbalance. Finally, it employs a CondConv-driven Boundary Refinement Module (BRM) for conditional, adaptive edge optimization. Evaluated on BSDS500 (ODS=0.829), NYUD-V2 (ODS=0.768), and BIPED (ODS=0.903), LUS-Net achieves state-of-the-art performance, significantly improving edge sharpness and structural fidelity compared to existing methods.

Technology Category

• Computer Vision: Learning & Optimization for CV
• Machine Learning: Learning on the Edge & Model Compression
• Search and Optimization: Learning to Search

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Cloud, edge and content delivery systems for the Web
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Edge detection is a fundamental task in computer vision. It has made great progress under the development of deep convolutional neural networks (DCNNs), some of which have achieved a beyond human-level performance. However, recent top-performing edge detection methods tend to generate thick and noisy edge lines. In this work, we solve this problem from two aspects: (1) the lack of prior knowledge regarding image edges, and (2) the issue of imbalanced pixel distribution. We propose a second-order derivative-based multi-scale contextual enhancement module (SDMCM) to help the model locate true edge pixels accurately by introducing the edge prior knowledge. We also construct a hybrid focal loss function (HFL) to alleviate the imbalanced distribution issue. In addition, we employ the conditionally parameterized convolution (CondConv) to develop a novel boundary refinement module (BRM), which can further refine the final output edge maps. In the end, we propose a U-shape network named LUS-Net which is based on the SDMCM and BRM for crisp edge detection. We perform extensive experiments on three standard benchmarks, and the experiment results illustrate that our method can predict crisp and clean edge maps and achieves state-of-the-art performance on the BSDS500 dataset (ODS=0.829), NYUD-V2 dataset (ODS=0.768), and BIPED dataset (ODS=0.903).

Problem

• Image Processing
• Edge Detection
• Mathematical Representation

Innovation

• SDMCM Module
• HFL Calculation
• CondConv Technology