Google has announced the release of Live Transcribe – a free Android service that provides real-time continuous transcription using speech recognition and Google Cloud.
Live Transcribe makes real-world conversations more accessible by bringing the power of automatic captioning into everyday, conversational use. The system uses Google’s advanced Automatic Speech Recognition (ASR) technology (which is already part of several Google projects: automated captions in Youtube, presentations in Slides and also phone calls) to detect audible languages and convert them into text.
Over 5% of the world’s population – or 466 million people are hearing-impaired or have disabling hearing loss, according to the World Health Organization. Services like LiveTranscribe can severely change the life of these people and help them participate in unrehearsed conversations as well as social occasions.
Live Transcribe supports over 70 languages and more than 80% of the world’s population. It provides captions conversations in real-time and it is integrated with Android. The service can be launched with a single tap from within any Android application. It is expected that LiveTranscribe will improve greatly the accessibility and help deaf and hearing-impaired people in their everyday life.
To build a transcribe service such as LiveTranscribe, Google implemented an on-device neural network-based speech detector, built on their previous work in automatic speech recognition.
According to Google, LiveTranscribe combines the results of extensive user experience (UX) research with seamless and sustainable connectivity to speech processing servers. The engineering team behind LiveTranscribe had to overcome a number of challenges among which: compute-intensive models, expensive access to connectivity, user experience etc.
The service is available on Play Store and is pre-installed on all Google Pixel 3 devices with the latest update. More about LiveTranscribe can be read at the official blog post.