Open Speech Tools Are Available and Gaining in Sophistication

by Ostatic Staff - Oct. 09, 2014

Today, with our ubiquitous mobile devices, many of us are used to talking to our gadgets to search for answers, get directions and more. This trend will only ramp up as smartwatches designed to respond to voice commands proliferate.

Not everyone knows, though, that there are some very notable open source projects focused on speech and speech recognition. These may play a bigger role in technology development as the age of reliable speech recognition is upon us. Here are a few open source speech projects to know about.

Wired U.K. has just done an interesting report on JuliusJS -- an open source tool that lets software developers build voice-controlled applications for web browsers. Its developers want to usher in a whole new class of Siri-like apps for the desktop. According to Wired's report:

"JuliusJS [has] a code library that runs inside the browser, and because it uses common web standards, it can do its thing on practically any machine. Using the library, developers could build tools for navigating from website to website, add voice controls into games, or do things no one has thought of yet."

One of the most robust open source speech recognition solutions comes from Carnegie Mellon University. It's called Sphinx, and we covered it here. You can use Sphinx for straight speech recognition, or integrate it with applications. To find out more about Sphinx, check out its hosted pages.

And, of course, we've been covering the advancement of voice and speech recognition features in browsers such as Google Chrome.  The latest versions of Android include responsiveness to voice commands, and there is one other interesting thing to take note of about Google in this area: Google recently hired futurist and tech pundit Ray Kurzweil, who is one of the world's leading experts on speech recognition and the pattern recognition science behind it. Kurzweil also specializes in text-to-speech technology, and has brought products to market based on it.

For simple commands, speech recognition science is now good enough to use, after years of being undependable.  Look for open source tools to help contribute to the advancement of these technologies.