Mechanisms for Hands-Free Point of Care Data Entry in Combat Environments

Based on the success of our Phase I research and demos, we further propose our novel and promising solution for hands- and eyes-free voice interface for point-of-care data entry for Phase II research. The solution consists of the following major modules: a recently-developed noise-reduction microphone, an automatic speech recognizer, a text-to-speech synthesizer, a set of medical grammar, and a dialogue manager. To make the developed system useful in combat environments, we emphasize noise robustness and user-friendly dialogue management. Our unique new research results and algorithms in noise reduction, auditory-based feature extraction, and discriminative training will be implemented to ensure the robustness; new development language and tools for dialogue management will be used in the system for fast and convenient development. Every effort has been made to minimize the project risk. In the end of this project, we will deliver a PDA with pre-installed software package developed from this project. Video demos and documents will be presented and submitted as well.