Google on Wednesday introduced a new artificial intelligence-based feature, called Translatotron, that enables users to instantly translate speech from one language to another.
The offline, end-to-end speech-to-speech translation model bypasses the typical two-step translation process, in which speech is first converted to text, which is then translated by a text-to-text model and finally turned back into speech.
With the new model, users can sidestep the time spent in the middle step, said Google, enabling .
