An Efficient Methodology for the Development and Evaluation of Advanced Embedded AI Technologies: FY23 Information, Computation, and Exploitation Line Supported Program
Abstract
This document provides a high-level overview of a research and development methodology which leverages a comprehensive framework to build and support an environment for efficient evaluation and design of AI technologies targeting embedded applications. More specifically, by utilizing containers, standard instrumentation utilities, and interface abstractions, it allows users to quickly develop and integrate targeted application test suites in order to jointly evaluate both embedded hardware processors and AI algorithms in the context of any application-of-interest.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 24, 2024
- Accession Number
- AD1222744
Entities
People
- Vitaliy Gleyzer
Organizations
- Massachusetts Institute of Technology