CallTran: Voice Translation for End-to-End Communication over the Internet
M.D Rahmy Athas, P. Pirapuraj · 2024
The importance of language translation becomes evident when two individuals, each speaking a different mother tongue and lacking a common language, seek to communicate effectively. This article presents the development and evaluation of a mobile application for voice-to-voice translation over the Internet. The application employs three main technologies: speech recognition, machine translation, and speech synthesis. Google APIs (speech-to-text API, text-to-text-translator API, and text-to-speech synthesizer API) were used to implement the system. The evaluation showed that the system achieved an overall accuracy of 85% in recognizing and translating speech input from users. However, the accuracy varied across different languages. The system was also found to be effective in facilitating communication between users who speak different languages. The limitations of the system were identified in its performance in noisy or crowded environments and the handling of regional accents and dialects. Overall, the developed system has the potential to bridge language barriers and facilitate communication among people speaking different languages.