Work

Texture- and Shape-Based Adversarial Attacks for Overhead Image Vehicle Detection

Year: 2025

Type: conference-paper

Abstract: Detecting vehicles in aerial images is difficult due to complex backgrounds, small object sizes, shadows, and occlusions. Although recent deep learning advancements have improved object detection, these models remain susceptible to adversarial attacks (AAs), challenging their reliability. Traditional AA strategies often ignore practical implementation constraints. Our work proposes realistic and practical constraints on texture (lowering resolution, limiting modified areas, and color ranges) and analyzes the impact of shape modifications on attack performance. We conducted extensive experiments with three object detector architectures, demonstrating the performance-practicality trade-off: more practical modifications tend to be less effective, and vice versa. We release both code and data to support reproducibility at https://github.com/humansensinglab/texture-shape-adversarial-attacks.

Source: Proceedings - International Conference on Image Processing

DOI: 10.1109/icip55913.2025.11084759

Authors: Mikael Yeghiazaryan, Sai Abhishek Siddhartha Namburu, Emily Kim, Stanislav Panev, Celso De Melo, Fernando De La Torre, Jessica K. Hodgins

Institutions: Carnegie Mellon University, DEVCOM Army Research Laboratory, United States Army Combat Capabilities Development Command, United States Army

Language: English

Topic: Adversarial Robustness in Machine Learning

Subfield: Artificial Intelligence

Field: Computer Science

Domain: Physical Sciences

Open Access status: green

Funder: Arm

OpenAlex ID: W4413278463

FWCI:

Cites: 12

Cited by:

Related to: 10