Recognition Framework for Efficient In-context Performance Measurement (REVOICE)

This Navy topic seeks to address the current lack of effective and efficient methods to integrate voice communication data for context-specific training performance measurement.  KBSI proposes to research, design, and demonstrate a Reliable Voice Recognition Framework for Efficient In-context Performance Measurement (REVOICE) that targets this challenge head-on. KBSI’s proposed REVOICE solution will provide several innovative capabilities including - intelligent speech recognition with smart voice profiling and noise filtering; natural language processing engine with advanced semantic analysis; context-specific training performance measurement that integrates voice data; an ‘instructor in a box’ to accelerate BREVITY term learning, adaptive training performance measurement, after action review; dynamic visualizations for enhanced situational awareness; and an open architecture framework to enable seamless integration with Naval training systems. Innovations include: (i) a reliable speech-to-text conversion technology with noise-robust word recognition for BREVITY-based communications, (ii) intelligent voice profiling to keep track of commands by source entity, (iii) in-context semantic interpretation of voice commands and automated mapping to the training scenario timeline, (iv) dynamic visualizations that integrate voice communication with tactical execution data, (v)  automated generation of After Action Review (AAR) reports, and (vi) a plug and play architecture for rapid integration with warfighter training systems.