Ever gone to a meeting only to end up in a heated discussion? You can always go over what was stated in the minutes, but unless you had an audio recording device with you, your mental recollection is all you have left to refer to. That won’t be the case if Japan’s National Institute of Advanced Industrial Science and Technology (AIST) has its way – they have managed to develop a system which is smart enough to determine who is speaking at that moment, recording every single word while producing proceedings that visualize the entire meeting structure. Part of the system includes a camera and microphone that records both audio and video for an analysis later, so speakers do not need to wear a microphone. Some of the components include :-
- An input device combining an 8-directional microphone array and an omni-directional camera
- Software to localize and separate a sound source, recognize different voices and so on
- The “MArcBrowser” to browse the multimedia proceedings
At long last, we will soon be able to reconstruct a meeting via these recordings instead of relying on printed minutes that might have missed out on a word or two.