An anonymous reader quotes TechCrunch: Microsoft announced today that its conversational speech recognition system has reached a 5.1% error rate, its lowest so far. This surpasses the 5.9% error rate reached last year by a group of researchers from Microsoft Artificial Intelligence and Research and puts its accuracy on par with professional human transcribers who have advantages like the ability to listen to text several times. Both studies transcribed recordings from the Switchboard corpus, a collection of about 2, 400 telephone conversations that have been used by researchers to test speech recognition systems since the early 1990s. The new study was performed by a group of researchers at Microsoft AI and Research with the goal of achieving the same level of accuracy as a group of human transcribers who were able to listen to what they were transcribing several times, access its conversational context and work with other transcribers. Read more of this story at Slashdot.
Read more here:
Microsoft Speech Recognition Now As Accurate As Professional Transcribers
sombragris writes: July 17 marked the 24th anniversary of Slackware Linux, the oldest GNU/Linux still in active development, being created in 1993 by Patrick Volkerding, who still serves as its BDFL. Version 14.2 was launched last year, and the development version (Slackware-current) currently offers kernel 4.9.38, gcc 7.1, glibc 2.25, mesa 17.1.5, and KDE and Xfce as official desktops, with many others available as 3rd party packages. Slackware is also among the Linux distributions which have not adopted systemd as its init system; instead, it uses a modified BSD init which is quite simple and effective. Slackware is known to be a solid, stable and fast setup, with easy defaults which is appreciated by many Linux users worldwide. Phoronix has a small writeup noting the anniversary and there’s also a nice reddit thread. Read more of this story at Slashdot.