How Interview Mode Elevates the Local AI Translator

Interview Mode Overview

Many users ask me, ‘What’s Interview Mode?’ in the Audio-2-Audio tab of the Local AI Translator.

The tab translates audio from one language to another in a single workflow, using a cascade of speech-to-text, translation, and text-to-speech.

First, it is not state-of-the-art speech-to-speech. The system uses a cascade approach to handle the work.

Two speakers are involved—ideally near a mic. Speaker A speaks, and the system translates it and voices it in Speaker B’s language.

Speaker B replies, and their reply is translated and voiced in Speaker A’s language.

On my Asus Zenbook 14 OLED this is a bit ponderous, yet it works as a rough workaround when you need to talk about a subject with someone who does not speak your language.

The big plus is that everything runs in local mode without Internet access.

There are current over a hundred voices spanning 35 languages, and this project is on the up. Expect more languages in the coming years.