Modular Design of a Front-End and Back-End Speech-to-Speech Translation Application for Psychiatric Treatment of Refugees

• Published in: 2023 IEEE Global Humanitarian Technology Conference (GHTC) pp. 128 - 131
• Main Authors: Ugan, Enes Yavuz, Mediani, Mohammed, Al Jawabra, Omar, Khader, Aya, Liu, Yining, Waibel, Alexander
• Format: Conference Proceeding
• Language: English
• Published: IEEE 12.10.2023
• Subjects: application; artificial intelligence; Data privacy; Faces; machine translation; Mental health; Sociology; speech recognition; speech synthesis; speech-to-speech system; Statistics; Task analysis
• ISSN: 2473-5728
• DOI: 10.1109/GHTC56179.2023.10354809

One of the inevitable impacts happening in areas with political conflicts is the significant influx of displaced individuals. The psychological consequences on individuals enduring such events are profound. Therefore, the imperative of providing adequate mental health care to refugees coming from conflict areas becomes apparent. However, providing this necessary care faces two obstacles. On the one hand, not all this target population is expected to have an acceptable level of proficiency of the hosting country's local language. On the other hand, finding enough number of suitable interpreters is a very challenging task. Moreover, even when the availability of the human interpreters is no problem, the refugees may hesitate to share their experiences with interpreters due to the associated stigma. To address these challenges and enhance mental health care for refugees, we propose the design of a modular front-end and back-end Speech-to-Speech translation system, with a focus on safeguarding patient data privacy. As our system is Speech-to-Speech, it also enables dialogue with dyslexic people and removes barriers for their treatment as well.