Microsoft has announced another feature deprecation today as it has killed off Speech Recognition and replaced it with Voice Access. The former was added back in the Windows Vista days.
Speech recognition RSS
From leveraging AI in its productivity suite to exploring the possibilities of techno-humanism, Microsoft continues venturing into the AI era as it embraces its power in its products and services.
Microsoft has confirmed speech recognition issues in Windows 11 version 22H2. According to the company, users with specific languages might experience problems with voice input, handwriting, and more.
Major companies like Microsoft, Apple, and Meta, among others, are collaborating with the University of Illinois under the Speech Accessibility Project to improve speech recognition facilities.
Mozilla has released its latest Common Voice dataset and for the first time, it contains 20,000 hours of content, which is almost double the number of hours that it had one year ago.
The Columbia University's computer science department has developed a new algorithm that can be used to stop microphones and automatic speech recognition systems from spying on the public.
By harmonizing modules pertaining to different media, Nvidia Jarvis will help build context for accurately predicting and generating responses in conversation-based AI applications.
Google Slides has gained a new capability meant to transcribe a spoken presentation in real time by accessing your computer's microphone to detect what you are saying. It will roll out next week.
Microsoft has once again achieved human-like speech recognition, after IBM set a new threshold for it. With a 5.1 percent word error rate, Microsoft is now the front runner in the industry.
Originally developed as a prototype during a company hackathon, Dictate promises to convert speech to text using the same "state of the art speech recognition" technologies used by Cortana.
IBM has announced that it has achieved the lowest word error rate in speech recognition, at 5.5%, beating Microsoft which managed 5.9% a few months back. This is still not equivalent to human parity.
The team over at Microsoft that works on speech recognition software have bested their own efforts, making history in the process, by achieving human-level understanding with speech recognition.
Microsoft has announced that it has achieved the lowest word error rate (WER) in speech recognition in the industry, which bodes well for its AI solutions, Cortana and Skype Translator.
Cortana will now be able to operate in more languages if you have the Windows Phone Developer Preview installed, with support added for the digital assistant in France, Germany, Italy and Spain.
Microsoft has announced this week that apps made for Windows 8.1 can now enable speech recognition features thanks to the new Bing Speech Recognition Control.
