The Effect of Data Augmentation in Deep Learning with Drone Object Detection
Ariel Yonatan Alin(1), Kusrini Kusrini(2*), Kumara Ari Yuana(3)
(1) Magister of Informatics Engineering, Universitas Amikom Yogyakarta
(2) Dept. of Information Technology, Universitas Amikom Yogyakarta,
(3) Dept. of Information Technology, Universitas Amikom Yogyakarta,
(*) Corresponding Author
Abstract
Drone object detection is one of the main applications of image processing technology and pattern recognition using deep learning. However, the limited drone image data that can be accessed for training detection algorithms is a challenge developing drone object detection technology. Therefore, many studies have been conducted to increase the amount of drone image data using data augmentation techniques. This study aims to evaluate the effect of data augmentation on deep learning accuracy in drone object detection using the YOLOv5 algorithm. The methods used in this research include collecting drone image data, augmenting data with rotate, crop, and cutout, training the YOLOv5 algorithm with and without data augmentation, as well as testing and analyzing training results.
The results of the study show that data augmentation can't improve the accuracy of the YOLOv5 algorithm in drone object detection. Evidenced by the decreasing value of precision and mAP@0.5 and the relatively constant value of recall and F-1 score. This is caused by too much augmentation, which can cause a loss of important information in the data and cause noise or distortion.
Keywords
Full Text:
PDFReferences
[1] L. Kapustina, N. Izakova, E. Makovkina, and M. Khmelkov, “The global drone market: main development trends,” SHS Web of Conferences, vol. 129, p. 11004, 2021, doi: 10.1051/shsconf/202112911004.
[2] Volunteers, “Drone Laws in Indonesia [Updated March 29, 2022],” 2022. https://drone-laws.com/drone-laws-in-indonesia/#Agencies_Responsible_for_regulating_drones_in_the_Republic_of_Indonesia (accessed May 30, 2022).
[3] D. Jones, “Small Remotely Piloted Aircraft Systems (drones),” 2017, Accessed: Jul. 28, 2023. [Online]. Available: http://www.nationalarchives.gov.uk/doc/open-government-licence/version/3/
[4] G. Lykou, D. Moustakas, and D. Gritzalis, “Defending airports from UAS: A survey on cyber‐attacks and counter‐drone sensing technologies,” Sensors (Switzerland), vol. 20, no. 12. MDPI AG, pp. 1–35, Jun. 01, 2020. doi:10.3390/s20123537.
[5] S. Singha and B. Aydin, “Automated drone detection using YOLOv4,” Drones, vol. 5, no. 3, 2021, doi:10.3390/drones5030095.
[6] C. Shorten and T. M. Khoshgoftaar, “A survey on Image Data Augmentation for Deep Learning,” J Big Data, vol. 6, no. 1, Dec. 2019, doi:10.1186/s40537-019-0197-0.
[7] Y. Zhou, S. Wen, D. Wang, J. Mu, and I. Richard, “Object detection in autonomous driving scenarios based on an improved faster-RCNN,” Applied Sciences (Switzerland), vol. 11, no. 24, Dec. 2021, doi: 10.3390/app112411630.
[8] W. Wu et al., “Application of local fully Convolutional Neural Network combined with YOLO v5 algorithm in small target detection of remote sensing image,” PLoS One, vol. 16, no. 10 October, Oct. 2021, doi: 10.1371/journal.pone.0259283.
[9] L. Jiao et al., “A survey of deep learning-based object detection,” IEEE Access, vol. 7, pp. 128837–128868, 2019, doi:10.1109/ACCESS.2019.2939201.
[10] A. Coluccia et al., “Drone vs. Bird detection: Deep learning algorithms and results from a grand challenge,” Sensors, vol. 21, no. 8, pp. 1–27, 2021, doi:10.3390/s21082824.
[11] J. Shijie, W. Ping, J. Peiyi, and H. Siping, “Research on Data Augmentation for Image Classification Based on Convolution Neural Networks.”
[12] S. O’gara and K. Mcguinness, “Comparing Data Augmentation Strategies for Deep Image Comparing Data Augmentation Strategies for Deep Image Classification Classification Comparing Data Augmentation Strategies for Deep Image Classification”, doi: 10.21427/148b-ar75.
[13] S. Park, S. Baek Lee, and J. Park, “Data augmentation method for improving the accuracy of human pose estimation with cropped images,” Pattern Recognit Lett, vol. 136, pp. 244–250, Aug. 2020, doi: 10.1016/j.patrec.2020.06.015.
[14] A. Buslaev, V. I. Iglovikov, E. Khvedchenya, A. Parinov, M. Druzhinin, and A. A. Kalinin, “Albumentations: Fast and flexible image augmentations,” Information (Switzerland), vol. 11, no. 2, Feb. 2020, doi: 10.3390/info11020125.
[15] T. DeVries and G. W. Taylor, “Improved Regularization of Convolutional Neural Networks with Cutout,” Aug. 2017, [Online]. Available: http://arxiv.org/abs/1708.04552
[16] D. Ho, E. Liang, and R. Liaw, “1000x Faster Data Augmentation – The Berkeley Artificial Intelligence Research Blog,” Jun. 07, 2019. https://bair.berkeley.edu/blog/2019/06/07/data_aug/ (accessed Apr. 24, 2023).
[17] G. Jocher et al., “ultralytics/yolov5: v4.0 - nn.SiLU() activations, Weights & Biases logging, PyTorch Hub integration,” Jan. 2021, doi: 10.5281/ZENODO.4418161.
[18] Z. Chen et al., “Plant Disease Recognition Model Based on Improved YOLOv5,” Agronomy, vol. 12, no. 2, 2022, doi: 10.3390/agronomy12020365.
[19] A. Biswal, “What is TensorFlow 2.0 [The Best Guide to Understand TensorFlow 2.0],” Simplilearn, Mar. 28, 2022. https://www.simplilearn.com/tutorials/deep-learning-tutorial/tensorflow-2 (accessed Dec. 05, 2022).
[20] S. Theophilus, “Roboflow: Converting Annotations for Object Detection | by Samuel Theophilus | Analytics Vidhya | Medium,” Medium, Sep. 03, 2021. https://medium.com/analytics-vidhya/converting-annotations-for-object-detection-using-roboflow-5d0760bd5871 (accessed Dec. 05, 2022).
DOI: https://doi.org/10.22146/ijccs.84785
Article Metrics
Abstract views : 2438 | views : 1856Refbacks
- There are currently no refbacks.
Copyright (c) 2023 IJCCS (Indonesian Journal of Computing and Cybernetics Systems)
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
View My Stats1