No more voice notes: WhatsApp works on a function to be able to transcribe them and thus convert them to text
The voice notes on WhatsApp they are a fast and above all comfortable method of sending a message. However, it is not always the same for the receiver, which often cannot hear the audio for different reasons. WhatsApp seems to have a solution for this, it is working on the possibility of transcribe voice memos to text in order to read them without having to listen to the audio.
As WABetaInfo, which has access to the development versions of the app, has learned, the service is testing a new feature for transcribing audios. Through it, it will recognize the words of an audio and automatically transcribe them to text to read the voice memo without listening to it.
At the moment some of the improvements that voice memos have in WhatsApp they have been, for example, to be able to change the playback speed. When it comes to transcribing them, it is currently also possible to do it using third-party apps.
Transcribe audios, without Facebook listening
One of the first questions that may arise with this is how Facebook will treat the data. Audio-to-text transcriptionIt requires analyzing the audio by a recognition engine in order to transcribe it. This implies … that Facebook reads the message? Although the company has an AI trained for this, it will not use it this time.
As we can see in the screenshots, the recognition is carried out using the own tools that the device already brings. In this case we see that an iPhone is used, so a message appears indicating that Apple’s voice recognition is used. The message also warns that the data can be sent to Apple to improve the system. This message is typical of iOS and other apps already use it.
Once the user accepts this condition, a new section will appear in the app where the transcribed messages are and they can be read. It will also be possible to jump to specific parts of the message, since the transcribed text is divided into small parts marked in time.
This feature certainly can be useful when reading voice memos in environments where they cannot be heard. For example in a public space where we do not want others to hear the voice memo or in a place where the noise is too loud to hear the voice memo clearly. It is also extremely useful, of course, for the hearing impaired.
At the moment functionality is under development, according to WABetaInfo. Therefore, we will have to wait to see it implemented in the app or even in the public beta of the app. We also have to see if it really arrives and the project is not abandoned, although being a relatively easy technology to implement and already present in mobile devices, it is more than likely that it will arrive.
One last aspect to highlight is which languages the service will recognize. In this case, English appears in the screenshots, but surely it is also in Spanish and many other languages from day one. Reason? It uses Apple technology (and probably that of Android manufacturers), which already works with dozens of different languages and dialects.
Via | WABetaInfo