Skip to content
DSML.KZ
Publications

Talk · International Conference on Image Processing Theory, Tools and Applications · October 13, 2025

A Frequency-Augmented CBAM Attention-Aided YOLOv8 for Object Detection

Sayani Dhar, Sourajit Maity, Asfak Ali, Dmitrii Kaplun, Daria Sidorina, Ram Sarkar

Dmitry Kaplun
Open via DOI

Abstract

This research proposes an improved object detection framework by integrating a Frequency-Augmented Convolutional Block Attention Module (FCBAM) into the YOLOv8 (You Only Look Once version 8) architecture. By incorporating frequency domain information, FCBAM enhances both spatial and channel attention mechanisms, enabling the model to focus more effectively on informative regions within an image. The attention module is embedded in a critical component of YOLOv8-Head to analyze its impact across the network. The effectiveness of the proposed approach is evaluated using three diverse domain-specific datasets: ADCOS for aerial aircraft detection, BRSTD for traffic and business sign recognition, and IRUVD for vehicle detection in Indian road conditions. Experimental results demonstrate that the FCBAM-enhanced YOLOv8 consistently outperforms the baseline, achieving mean Average Precision (mAP@0.5) scores of$\mathbf{0. 9 9 1}$on BRSTD,$\mathbf{0. 9 5 3}$on ADCOS, and 0.945 on IRUVD.

Advanced Neural Network ApplicationsVisual Attention and Saliency DetectionDomain Adaptation and Few-Shot LearningYOLOv8CBAM attentionobject detection

More by Dmitry Kaplun