 |
International Society of Science and Applied Technologies |
| Exploring Convolutional Neural Networks for NASA Mars Image Data | ||||
| Author | Kehan Gao
|
|||
| Co-Author(s) | Sarah Tasneem; Taghi M. Khoshgoftaar
|
|||
| Abstract | Deep learning, particularly through the use of Convolutional Neural Networks (CNNs), has played a significant role in computer vision, including image recognition and classification. This study investigates the performance of two essential CNN architectures, AlexNet and ResNet, in addressing class imbalances within image datasets. Specifically, we assess their efficacy using a NASA Mars image dataset characterized by varying levels of class imbalance. Our analysis focuses on comparing how each architecture performs under different imbalance scenarios, ranging from mild to severe. The empirical results demonstrate that while both models effectively manage mild imbalances, AlexNet’s performance significantly declines with increased imbalance severity, unlike ResNet, which demonstrates considerable robustness even in highly imbalanced conditions.
|
|||
| Keywords | Deep Learning, Convolutional Neural Networks, AlexNet, ResNet, Image Classification, Class Imbalance | |||
| Article #: RQD2024-159 | ||||
Proceedings of 29th ISSAT International Conference on Reliability & Quality in Design |