Developing Multi-Voice Speech Recognition Confidence Measures and Applying Them to AHLTA-Mobile

Abstract

While developing a documentation solution for medics in the field, confidence measures have been incorporated to improve command rejection and minimize enrollment. The new enrollment algorithms makes speaker-dependent recognition competitive with speaker-independent recognition.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
May 01, 2011
Accession Number
ADA547479

Entities

People

  • Gregory J. Gadbois

Tags

Communities of Interest

  • Biomedical
  • C4I
  • Human Systems
  • Materials and Manufacturing Processes

DTIC Thesaurus Topics

  • Accuracy
  • Algorithms
  • Automated Speech Recognition
  • Cardiovascular Physiological Phenomena
  • Command And Control
  • Computer Science
  • Health Services
  • Measurement
  • Military Hospitals
  • Operating Systems
  • Patient Care
  • Recognition
  • Rejection
  • Signal Processing
  • Supervised Machine Learning
  • User Interface
  • Vital Signs

Readers

  • Military Training and Readiness Simulation
  • Speech Processing/Speech Recognition.
  • Systems Analysis and Design

Technology Areas

  • AI & ML
  • AI & ML - Bayesian Inference
  • AI & ML - DoD AI Strategy