PHITE: Portable High-performance Inference at the Tactical Edge

Abstract

Problem: Todays AI software is computationally expensive and requires extensive knowledge, skill, and effort to adopt on low-power devices at the tactical edge. Solution: Develop an open-source library of machine learning (ML) algorithms optimized for low-power (100s mW10s W) embedded devices.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Nov 14, 2022
Accession Number
AD1183612

Entities

People

  • Jay Palat
  • Navya Chandra
  • Nicolai Tukanov
  • Oren Wright
  • Pankti Shah
  • Scott McMillan
  • Tze M. Low
  • Upasana Sridhar

Organizations

  • Carnegie Mellon University

Tags

Communities of Interest

  • Autonomy
  • Energy and Power Technologies

DTIC Thesaurus Topics

  • Algorithms
  • Anomaly Detection
  • Artificial Intelligence
  • Change Detection
  • Classification
  • Computers
  • Convolution
  • Department Of Defense
  • Detection
  • Detectors
  • Engineering
  • Image Classification
  • Learning
  • Machine Learning
  • Materials
  • Neural Networks
  • Software Development
  • Target Acquisition
  • Target Recognition
  • Training
  • Universities
  • Validation

Fields of Study

  • Computer science

Readers

  • Database Systems and Applications
  • Integrated Circuit Design and Technology.
  • Military Training and Readiness Simulation

Technology Areas

  • AI & ML
  • AI & ML - DoD AI Strategy