DOI: 10.5176/2301-394X_ACE19.581
Authors:Gregory Thomas
Abstract: Audible communication is a necessary and vital tool in healthcare environments for patient comfort, care planning, and health literacy. Patients use speech to navigate through appointments and treatments and to seek answers concerning their illnesses. Doctors use it to quickly and accurately communicate with clinical staff. Nurses use it to interact with patients and for myriad other needs. In such speech-dependent settings, speech-application technologies, such as Amazon Echo, Apple’s Siri, Samsung’s Bixby, and Google Home, can provide significant benefits by improving efficiencies and streamlining communications within a major medical environment. This paper describes a proposal to investigate the adoption by hospital environments of speech-application technologies tailored to support healthcare-specific needs. Specifically, we will identify family and staff communication needs best suited for Automatic Speech Recognition, including ASR-device support as well as Natural language processing (NLP). NLP technologies are used for information extraction, automatic speech recognition, machine translation, and dialogue systems, with a focus on providing access and proximity to patients. Using aspects of these two languagelearning and application disciplines in the form of voice-controlled devices, we will implement and evaluate speech-recognition engagement in hospital settings.
Keywords: Amazon Echo, automatic speech recognition, automatic speech, voice recognition, DeepSpeech, voice-based machine translation, voice activity detection (VAD), natural language processing, NLP, ASR, language learning, machine learning, voice-controlled device, voice-activated technology, application adaptation, speech recognition API, keyword spotting, speaker diarization
Updating...