Abstract
In order to separate multiple voices, various Speech Separations has been examined using microphone array. But, they are on the assumption that multiple humans are static. So moving multiple humans are not considered. We have investigated Speech Separation considering moving humans. To separate moving multiple voices, we need to measure positions of multiple humans at the same time as record voices. We have focused on depth image sensor to measure positions of multiple humans. At first, measuring positions of multiple humans using depth image sensor. Then we try Speech Separation based on multiple human positions. We experiment on Speech Separation with microphone array and depth image sensor.
