Recently Google has come up with a new technology to power its voice search,
Google Now, which the company says will make it more accurate and even faster. This new technology allows Google voice search recognition to have a better neural network acoustic models using Connectionist Temporal Classification (CTC) and some sequence training techniques . These models are some kind of special extension of recurrent neural networks (RNNs) that are more precise and accurate, especially in noisy environments, and they are amazingly fast!
Read more: Turn Iphone Into a PS Vita
Back in 2012, Google switched from Gaussian Mixture Model (GMM) to Deep Neural Networks (DNNs), which allowed the company to better assess and evaluate which sound a user was producing at that time, and delivered an increased Google voice speech recognition accuracy.
I used to think that ‘If it ain’t broke, don’t fix it’ but if there is room for Improvement then why not? Well it’s certainly seems to be a step closer to everyone’s dream to ask their computer a question as if it was a human being.