On Thursday, Google announced that it has improved its voice search tool, making it faster and more accurate.
Since 2012, Google voice search has used Deep Neural Networks, or DNNs, and the core technology used to model language sounds. Now, the company has taken the next step in speech recognition accuracy. "Today, we're happy to announce we built even better neural network acoustic models using Connectionist Temporal Classification (CTC) and sequence discriminative training techniques," explained Google in a post on its Google Research Blog. "These models are a special extension of recurrent neural networks (RNNs) that are more accurate, especially in noisy environments, and they are blazingly fast!" Compared to traditional speech recognizers, the new technology can predict phenomes faster and can differentiate between similar sounding letters while requiring lower computational resources to analyze sounds in real-time.
"We are happy to announce that our new acoustic models are now used for voice searches and commands in the Google app (on Android and iOS), and for dictation on Android devices," said Google. "In addition to requiring much lower computational resources, the new models are more accurate, robust to noise, and faster to respond to voice search queries - so give it a try, and happy (voice) searching!" Google is not the only one working to improve voice search; Apple, Microsoft, Facebook, and others are working to improve and add voice recognition features. Facebook recently acquired speech recognition company Wit.ai, but has yet to debut any new features with its technology.
Photo: © Google.