Sensitivity analysis of ResNet-based automatic target recognition performance using MuSES-generated EO/IR synthetic imagery
When
Where
Abstract: Machine learning algorithms have demonstrated state-of-the-art automated target recognition performance but require a large training set. In the case of electro-optical infrared (EO/IR) remote sensing, acquiring sufficient measured imagery can be difficult, but EO/IR scene simulation is a possible alternative. CoTherm, a co-simulation tool which operates MuSES in an automated fashion, is used to manipulate relevant target, background and sensor inputs to generate a library of radiance images. Various options affecting simulation run-time and output fidelity are considered and the trade-off between accuracy and compute time requirements is quantified using a measured imagery benchmark and ResNets for classification.
Zoom: https://arizona.zoom.us/j/87104438814
Following Matt Young’s talk will be several short talks:
Speaker: Eric Anderson (RTX, Senior Principal Engineer)
Title: Covariances and the Eigenvalue Decomposition
Time: 3:40-4:20
Speaker: Theodore Meissner, Program in Applied Mathematics, University of Arizona
Title: Sparse Optimization Techniques for Differential Equation Discovery
Time: 4:25-4:55
Speaker: Patrick Kano (RTX, Principal Engineer)
Title: Dempster-Shafer Evidential Theory for Optimal Contours in Talbot’s Method
Time: 5:00-5:40
5:40-5:50 Break
Speaker: Nitesh Shah (RTX, Senior Principal Engineer)
Title: Introduction to RTX Processing, Guidance, and Control Center
Time: 5:50-6:15
Speakers: Lexi Casarez (RTX, Campus Programs Recruiter), RTX Staff and UAZ Faculty
Title: Question and Answer Session
Time: 6:15-6:30
Speaker: Michael (Misha) Chertkov, Professor and Chair, Program in Applied Mathematics, University of Arizona
Title: Closing
Time: 6:30