Graphically Residual Attentive Network for tackling aerial image occlusion

Article Type

Research Article

Publication Title

Computers and Electrical Engineering

Abstract

Deep learning has rapidly advanced, enabling new applications such as object detection, text recognition, and occlusion handling. However, challenges remain in the detection of objects in complex environments such as aerial images where things like motion blur, low light, and significant occlusion occur. This paper addresses a similar challenge by introducing a novel supervised framework, the Graphically Residual Attentive Network (GRESIDAN). In the same model, GRESIDAN integrates three synergistic pipelines for object detection, occlusion detection, and occlusion removal. GRESIDAN uses a residually attentive block combining ResNet-18 and a multi-headed attention mechanism to improve feature extraction and detection accuracy in low-quality, occluded aerial images. A graphically attentive occlusion detection pipeline is implemented to handle occlusion, segment better, and mask out the occluder in the aerial image. The GRESIDAN model is validated on the COCO-2017 dataset and a custom private aerial object detection dataset, outperforming the state-of-the-art methods in handling occlusion and detecting objects. Our contributions provide a robust solution to the problem of detecting and handling occluded objects in aerial imagery, pushing the boundaries of automated visual recognition in challenging real-world scenarios. The code for public use in training and testing is available on GitHub.

DOI

10.1016/j.compeleceng.2025.110429

Publication Date

7-1-2025

Link to Full Text

Share

COinS