The difficulty in obtaining labeled data relevant to a given task is among the most common and well-known practical obstacles to applying deep learning techniques to new or even slightly modified domains. The data volumes required by the current generation of supervised learning algorithms typically far exceed what a human needs to learn and complete a given task. We investigate ways to expand a given labeled corpus of remote sensed imagery into a larger corpus using Generative Adversarial Networks (GANs). We then measure how these additional synthetic data affect supervised machine learning performance on an object detection task.<p> </p>Our data driven strategy is to train GANs to (1) generate synthetic segmentation masks and (2) generate plausible synthetic remote sensing imagery corresponding to these segmentation masks. Run sequentially, these GANs allow the generation of synthetic remote sensing imagery complete with segmentation labels. We apply this strategy to the data set from ISPRS' 2D Semantic Labeling Contest - Potsdam, with a follow on vehicle detection task. We find that in scenarios with limited training data, augmenting the available data with such synthetically generated data can improve detector performance.
Stand-off base and force protection surveillance measures primarily rely on electro-optic and thermal imaging technology. Atmospheric turbulence causes blur, distortion and intensity fluctuations that can severely degrade the image quality of these systems. This work explores the effects of turbulence image degradation on the performance of automatic facial recognition software and also looks at the potential benefit of turbulence mitigation algorithms. The goal of this work is to understand the feasibility of long-range facial recognition in degraded imaging conditions. In order to create a large enough database to match against, simulated imagery of different ranges and turbulence conditions were created using a horizontal view turbulence simulator and a subset of the Facial Recognition Technology (FERET) database. The simulated turbulence degraded imagery was then processed with facial recognition software and the results are compared against those from the pristine image set. Finally, the performance of the facial recognition software with turbulence mitigated imagery is also presented.
Conference Committee Involvement (2)
Applications of Machine Learning 2020
23 August 2020 | San Diego, California, United States
Applications of Machine Learning
13 August 2019 | San Diego, California, United States