Created an Accent-based Text-to-Speech Conversion System using HTML and JS, harnessing web-based technology to generate speech from written text while faithfully replicating the speaker’s accent. Achieved through the use of APIs to deliver access to 200 distinct accents and language models. Technologies used are HTML, CSS, JavaScript, BootStrap3, Web speech Utterance API.
