Re-Configurable Electro-Optical Device for Accelerating Deep-Network Training and Inference on Small Autonomous Platforms
Abstract
: Deep networks have demonstrated immense promise for enabling autonomy. However, the current paradigm of using stream-computing hardware, like graphics processing units (GPUs), for inferring network responses is ill suited for many types of robotic platforms. Itis especially troublesome for those platforms performing long-term missions with limited battery lifetimes. Going forward, it is crucial to retain the universality and flexibility of non-linear-inference capability of deep networks while reducing their overall power requirements. Doing so will help design a new generation of intelligent robotic systems for computing at the edge In this grant,we will develop a novel electro-optical device for analog, on-platform deep learning.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Apr 12, 2023
- Source ID
- N000142312363
Entities
People
- Sanjeev J. Koppal
Organizations
- Office of Naval Research
- United States Navy
- University of Florida