Since FTL has been developing real-time algorithms for sensing and analysis applications, machine learning and neural networks have progressed from academic anomalies, to exotic approaches for supercomputing, to veritable swiss-army-knife libraries applicable to a staggering variety of applications. Starting from classifiers for analyzing images based on thousands of training images, FTL is now applying machine learning to many problems in image analysis, 3D spatial awareness, situational prediction, and smart sensing. The links below provide summaries of projects in which FTL has applied machine learning to achieve cutting-edge system performance.