Speech and Translation Technologies

Abstract

mark seligman IntroductionCross-language communication in healthcare is urgently needed. Daily and nightly throughout the world, thousands of conversations are required between caregiversdoctors, nurses, administrators, volunteers, and othersand patients or family members with differing native languages.Chapter 1 describes and illustrates the exploding development of the relevant linguistic technologiesmachine translation (MT) of text, automatic speech recognition (ASR), and text-to-speech (TTS). The related infrastructurewireless communication, cloud computing, and mobile deviceshas also been developing apace. This chapter will shift focus to the combination and application of these technologies in the healthcare context, with special interest in speech translation.Given this impressive and accelerating progress, we'd expect various automatic translation and speech-enabled systems to be in widespread use by now; in fact, however, adoption remains slow. We'll examine the obstacles to adoption and directions for overcoming them in Section 2.2. In Section 2.3, we'll examine two major types of speech translation systems, concentrating on their respective approaches to the same obstacles. Section 2.4 will survey some healthcare-oriented communication systems, past and future. We'll conclude with an optimistic forecast for speech and translation applications in healthcare, tempered by due cautions. Obstacles to Adoption and Potential SolutionsOne key factor in the lagging adoption of linguistic technology in healthcare is the sheer difficulty of understanding the relevant technologies, and thus the natural hesitation to trust them. Accordingly, Chapter 1 aimed to bridge the 47