Voice recognition

While this tutorial has been all about text-to-speech software, at the other end of that spectrum is voice recognition. In a world driven by Alexa, Cortana and the like, it shouldn’t come as any surprise that voice recognition on Linux is also being actively pursued. Unfortunately, there doesn’t as yet exist a perfect solution to serve your every command.

The past decade has witnessed the rise, and sad demise, of various voice recognition solutions for Linux. The Wikipedia page on speech recognition software for Linux https://en.wikipedia.org/wiki/ Speech_recognition_software_for_linux has an incomplete list of various projects that are working towards this goal, and although not officially dead, many (see Roundup LXF188) of these applications have not witnessed a new release in several years.

One of most popular voice recognition systems today is the Mycroft artificial intelligence system, which is also popular for its use in various RPI projects. One factor responsible for its surge in popularity is cross-platform support, meaning you can also run it on a Linux desktop and even your Android phone.

For its own speech synthesis – the artificial production of human speech, such as what we have heard throughout this tutorial with espeak NG – Mycroft uses Mimic, which in turn is based on Flite, or Festival Lite, designed at Carnegie Mellon University.

Head over to https://mycroft.ai for the latest developments and more details on this ground-breaking project.