ME4824 Applications of Deep Learning for Military Systems

This course covers the applications of deep learning for military systems including satellite imagery processing and target detection, classification and aimpoint tracking for high energy laser beams.  The coverage begins with introduction to computer vision, machine learning, and "shallow" neural networks.  Then it expands to deep neural networks, Convolutional Neural Networks (CNN), network architecture design, training and validation, neural network databases, transfer learning, and CNN frameworks including TensorFlow. Laboratory sessions include network design, training, and inference for shallow and deep neural networks using Python programming language.  Tailoring various existing CNN network structures for transfer learning will be included as well for satellite imagery processing and target object detection, classification, and tracking.

Cross Listed Courses

Cross-listed with AE4824

Prerequisite

Knowledge of controls and basic programming or consent of instructor

Lecture Hours

3

Lab Hours

2

Course Learning Outcomes

The primary outcome of the course is for students to gain knowledge in the development and application of deep learning algorithms for image data analysis of military systems. At the completion of this course students will demonstrate competence in the following areas.

• Design, training, and inference of neural networks.
• Design, training, and inference of deep learning networks.
• Design, training, and inference of Convolutional Neural Networks (CNN).
• Transfer learning with training databases and CNN framework such as TensorFlow.
• Application of deep learning in satellite image object detection and classification.
• Application of deep learning in High Energy Laser target detection, classification, and tracking.