Google is a company characterized by having a large number of open source projects. But recently he has made some very interesting moves for the open-source community, and that is to open two of his big projects. These are related to Android, but both are quite different from each other. Now they will be available from GitHub to everyone for a few weeks ...
The first project we are talking about is the voice engine for Live Transcribe, an interesting tool for speech recognition and speech-to-text transcription for Android. It is based on machine learning algorithms to transform audio into subtitles in real time. If you think about it, it is not only important for the community, but also to improve the accessibility of Linux systems, since it is not an aspect of which we have to be extremely proud and much more could be done.
On the other hand, Google has also opened another of its software systems. And it is that of gesture tracking for Android. These types of projects are not very frequent among those that are already available open source, so it is good news. Having this code will allow developers to create better systems based on them. And, I repeat, it is something important to improve accessibility or certain aspects that are not too careful in open source operating systems, although in Android this aspect has been improved a lot ...
In this second released system, the gesture tracking system, it is built on a MediaPiple machine learning framework. This software combines three components of artificial intelligence: a palm detector, a model that returns 3D hand points, and a gesture recognizer. In this way, the experience in a wide variety of domains and technological platforms will be improved.
Github - speech recognition
Github - gesture recognition
Hopefully someone is interested in developing something with the live translator. Those of us who work with interviews and that kind of thing will thank you forever!