This demo showcases simultaneous speech translation from English into Chinese, Japanese, and German. The system translates incoming speech incrementally, providing both translated text (speech-to-text) and spoken output in the target language (speech-to-speech). The recordings below demonstrate these two output formats in action.

The English speech used as translation input comes from this original YouTube video.

Note: The demos were recorded in a real-world environment, with ambient noise and fluctuations in network connectivity.