AeroNet: Efficient YOLOv7 for Tiny-Object Detection in UAV Imagery
##plugins.themes.academic_pro.article.sidebar##
##plugins.themes.academic_pro.article.main##
Abstract
The detection of multiple tiny objects from diverse perspectives using unmanned aerial vehicles (UAVs) and onboard edge devices presents a significant challenge in computer vision. To address that, this study proposes AeroNet, a lightweight and efficient detection algorithm based on YOLOv7 (You Only Look Once version7).This algorithm features the LHGNet (Lightweight High-Performance GhostNet) backbone, an advanced feature extraction network that integrates depth-wise separable convolution and channel shuffle modules.These modules enable deeper exploration of network features, promoting the fusion of local detail information and channel characteristics. Additionally, this research introduces the LGS(Lightweight Gradient-Sensitive) bottleneck and LGSCSP(Lightweight Gradient-Sensitive Cross Stage Partial Network) fusion module in the neck to reduce computational complexity while maintaining accuracy. Structural modifications and adjusted feature map sizes further enhance detection accuracy. Evaluated on the SkyFusion dataset,this method demonstrated a 25.0% reduction in parameter count and a 12.8% increase in mAP (0.5) compared to YOLOv7. These results underscore the effectiveness of this proposed approach in improving detection accuracy and model efficiency through the proposed enhancements.
##plugins.themes.academic_pro.article.details##
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
References
- Bochkovskiy, A., Wang, C., and Liao, H. Y. M. 2020. Yolov4: Optimal speed and accuracy of object detection. arXiv preprint.
- Li, C., Li, L., Jiang, H., Weng, K., Geng, Y., and et al. 2022. Yolov6: A single-stage object detection framework for industrial applications. arXiv preprint.
- Li, C., Zhou, A., and Yao, A. 2022. Omni-dimensional dynamic convolution.
- Li, S., Yang, X., Xiao, L., Zhang, Y., and Wu, J. 2023. Real-time vehicle detection from uav aerial images based on improved yolov5. Multidisciplinary Digital Publishing Institute 23, 12 (June 16), 5634–5634. DOI: https://doi.org/10.3390/s23125634
- Liu, K., Peng, L., and Tang, S. 2023. Underwater object detection using tc-yolo with attention mechanisms. Sensors 23, 5, 2567. DOI: https://doi.org/10.3390/s23052567
- Liu, M., Wang, X., Zhou, A., Fu, X., Ma, Y., and Piao, C. H. 2020. Uav-yolo: Small object detection on unmanned aerial vehicle perspective. Multidisciplinary Digital Publishing Institute 20, 8, 2238–2238. DOI: https://doi.org/10.3390/s20082238
- Liu, Z., Mao, H., Wu, C.-Y., Feichtenhofer, C., Darrell, T., and Xie, S. 2022. A convnet for the 2020s. arXiv preprint. DOI: https://doi.org/10.1109/CVPR52688.2022.01167
- Shao, Z., Wu, W., Wang, Z., and et al. 2018. Seaships: A large-scale precisely annotated dataset for ship detection. IEEE Transactions on Multimedia 20, 20, 2593–2604. DOI: https://doi.org/10.1109/TMM.2018.2865686
- Wang, C.-Y., Bochkovskiy, A., and Liao, H.-Y. M. 2022. Yolov7: Trainable bag-of-freebies sets new state-of-the-art for realtime object detectors. arXiv preprint. DOI: https://doi.org/10.1109/CVPR52729.2023.00721
- Wen, G., Li, S., Liu, F., Luo, X., Er, M.-J., Mahmud, M., and Wu, T. 2023. Yolov5sca: A modified yolov5s network with coordinate attention for underwater target detection. Sensors 23, 7, 3367. DOI: https://doi.org/10.3390/s23073367
- Xu, Z., Shi, H., Li, N., Xiang, C., and Zhou, H. 2018. Vehicle detection under uav based on optimal dense yolo method. DOI: https://doi.org/10.1109/ICSAI.2018.8599403
- Zhang, S., Chai, L., and Jin, L. 2020. Vehicle detection in uav aerial images based on improved yolov3. DOI: https://doi.org/10.1109/ICNSC48988.2020.9238059