Freq-DP Net: A Dual-Branch Network for Fence Removal using Dual-Pixel and Fourier Priors

2026-02-15

AI Summary

This work addresses the challenging problem of fence occlusion removal from a single image, where existing methods either suffer from limited performance in static scenes or rely on multi-frame motion cues. To overcome these limitations, we propose Freq-DP Net, the first approach to leverage dual-pixel (DP) sensor data for single-image fence removal. Our method introduces an end-to-end dual-branch network: one branch explicitly models geometric priors using defocus disparity, while the other captures global structural priors of fences in the frequency domain via fast Fourier convolution (FFC). An attention mechanism adaptively fuses these complementary features to enable precise fence segmentation. Evaluated on a newly curated, diverse fence dataset, our approach significantly outperforms existing baselines and establishes a new state-of-the-art for single-image DP-based fence removal.

Technology Category

• Computer Vision: Other Foundations of Computer Vision
• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Search and Optimization: Learning to Search

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Security and Privacy: Large-scale security measurements
• Systems and Infrastructure for Web, Mobile and WoT: Federated Web and WoT systems, including distributed, federated and edge-based data processing

Abstract

Removing fence occlusions from single images is a challenging task that degrades visual quality and limits downstream computer vision applications. Existing methods often fail on static scenes or require motion cues from multiple frames. To overcome these limitations, we introduce the first framework to leverage dual-pixel (DP) sensors for this problem. We propose Freq-DP Net, a novel dual-branch network that fuses two complementary priors: a geometric prior from defocus disparity, modeled using an explicit cost volume, and a structural prior of the fence's global pattern, learned via Fast Fourier Convolution (FFC). An attention mechanism intelligently merges these cues for highly accurate fence segmentation. To validate our approach, we build and release a diverse benchmark with different fence varieties. Experiments demonstrate that our method significantly outperforms strong general-purpose baselines, establishing a new state-of-the-art for single-image, DP-based fence removal.

Problem

• fence removal
• occlusion removal
• single-image processing
• visual quality degradation
• computer vision applications

Innovation

• Dual-Pixel
• Fast Fourier Convolution
• Fence Removal
• Cost Volume
• Attention Mechanism