Computer Sciences

New AI boosts teamwork training

Researchers have developed a new artificial intelligence (AI) framework that is better than previous technologies at analyzing and categorizing dialogue between individuals, with the goal of improving team training technologies. ...

Engineering

'PrivacyMic': For a smart speaker that doesn't eavesdrop

Microphones are perhaps the most common electronic sensor in the world, with an estimated 320 million listening for our commands in the world's smart speakers. The trouble is that they're capable of hearing everything else, ...

Software

Microsoft adds transcription feature to Word

Microsoft launched a new transcription service for its online version of Word this week. Transcribe in Word allows users to record live conversations with multiple speakers or upload prerecorded audio or even video files.

Computer Sciences

Emotion recognition has a privacy problem—here's how to fix it

With devices listening everywhere you go, privacy concerns are endemic to advancing technology. Especially sensitive are different techniques powered by audio from your smartphones and speakers, putting consumers in a constant ...

Computer Sciences

How voice assistants follow inaudible commands

An attack against speech recognition systems with manipulated audio files used to work only via a data interface. Now, all it takes is playing the secret messages via loudspeakers.

Computer Sciences

Using math to blend musical notes seamlessly

In music, "portamento" is a term that's been used for hundreds of years, referring to the effect of gliding a note at one pitch into a note of a lower or higher pitch. But only instruments that can continuously vary in pitch—such ...

page 2 from 7