Deep Learning
Training Networks to Learn and Decide
UDRI researchers apply their expertise in deep learning in order to develop algorithms that train neural networks to learn unsupervised from collected data and make specific predictions. These algorithms then improve user awareness of available data, suggest areas for further investigation, and make connections between users and information to ensure analyses have more breadth and depth than a single user working alone. Our systems learn to understand how data products are developed, foster outside-the-box learning for users, and improve workflows.
Capabilities
- Deep learning architecture to identify illness such as lung cancer and pneumonia on medical images
- Deep learning algorithm that searches for markings on X-rays that indicate the presence of COVID-19
- Data analytics for finding activity in isolated environments with various, disparate sensors, accomplished in part through the use of the Open Standard for Unattended Sensors (OSUS), which UDRI developed in collaboration with the US Army Research Laboratory (ARL)
- Deep networks to detect and characterize Coordinate Measuring Machine (CMM) probes in order to improve quality assurance of machined parts
- Innovative systems that employ deep learning for predicting supply chain risks. These systems expose previously unidentified correlations, and, in turn, help assess new kinds of industrial risk.
- Video summarization by leveraging computer vision, audio processing, and natural language processing technologies. This capability fuses information from imagery and audio in order to infer delivery method, context, content, intent, and pedigree.