Mozilla partners with NVIDIA to transform Voice Recognition
With the advent of technology, we get to see new innovations with every passing day. Off late, voice-based applications, products and services ruled the world. On that note, NVIDIA invests $1.5 million in Mozilla Common voice that takes voice recognition to a different level. Seeing how the world is showing keen interest in the voice-enabled products, more and more organizations are coming up with ways that give a new dimension to voice recognition. This is exactly what NVIDIA has done.
Over the coming few years, it is highly likely that speech becomes a medium in which people could react with devices. Right from phones, laptops to digital assistants – we’ll see it all. Though we are at a stage where we get to see advanced application of voice recognition. But, today, the voice-enabled devices are not advanced to the extent that they can address all the languages of the world, various accents present and a range of speech patterns. Now, here is the catch. The coming few decades will take into account all of this and the days when we get to see voice-based applications and models being able to recognize speech, accents and patterns would not be too far.
The reason why Mozilla partners with NVIDIA and why the latter is investing a whopping $1.5 million in an initiative that is aimed at democratizing and diversifying voice technology development is to ensure that people from every corner of the world benefit from the massive technological shift, as discussed. This ambitious initiative of Mozilla Common voice will cater to this.
So far, we have seen machine learning algorithms being the base of voice data applications in almost all the major companies that you can think of. However, the challenge was when the objective was to develop high-quality speech recognition technologies and when one needs to differentiate between English speakers and the rest of the world.
Now enters ‘Common voice’. This was launched in 2017. This saw people donating their voices to a free, publicly available database. Following this, start-ups, researchers, and developers made use of these voices to train voice-enabled apps, products, and services. Today, this has gain immense popularity all across the globe to the extent that Common voice stands as the world’s largest multi-language public domain voice data set.
This boasts of more than 9,000 hours of voice data in 60 different languages. Everything, right from widely spoken languages like English to less spoken ones like Welsh and Kinyarwanda, this platform has got it all. The platform has set records in almost every possible aspect.
One of them is the number of people who have contributed to this project. Any guesses? Well, over 164,000 people!!!!!
Realizing the potential this platform holds, the coming years will see more of its implementation. Also, with the investment made by NVIDIA, the rising growth of Common voice’s data set would come as no big surprise. Additionally, this would open up opportunities for people who are looking for career opportunities.
In order to support the expansion, Common voice will now form a part of critical initiatives taken up by Mozilla Foundation. This is yet another step to make artificial intelligence more trustworthy.
