International Journal of Computer Techniques Volume 12 Issue 3 | Deep Learning Framework for Shipwreck and Airplane Classification in Sonar Images Using VGG19 and Siamese Networks
International Journal of Computer Techniques Volume 12 Issue 3 | Deep Learning Framework for Shipwreck and Airplane Classification in Sonar Images Using VGG19 and Siamese Networks
Sreeja Gundu Department of Electronics and Communication Engineering, Osmania University, Hyderabad, India gundu.sreeja@gmail.com
Abstract
This paper explores deep learning-based classification of sonar images into shipwrecks and airplanes using VGG19 and Siamese networks, with the VGG19 model achieving superior performance in underwater object detection.
Conclusion
The study validates the effectiveness of VGG19 for sonar image classification, demonstrating higher accuracy compared to the Siamese Network. Future enhancements will focus on dataset expansion, transformer-based models, and hybrid deep learning architectures.
References
F. Z. C. C. X. H. G. P. Xin Wen, βSide-Scan Sonar Underwater Target Detection,β IEEE Journal of Oceanic Engineering, 2024.
A. M. N. A. M. Tincy Thomas Chungath, βTransfer Learning-Based Deep Neural Networks for Underwater Sonar Image Classification,β IEEE Journal of Oceanic Engineering, 2024.
G. Koch et al., βSiamese Neural Networks for One-Shot Image Recognition,β ICML Deep Learning Workshop, 2015.
Post Comment