Subject of project/Short description (480 characters for publication purposes) Today, communication noises in mobile networks are often suppressed to improve speech intelligibility. But newer speech network generations can now transmit much higher quality signals, hence the potential to better exploit the information about a talker’s environmental context. New international ITU standards will soon be defined along that trend. The present project is truly innovative, and will exploit leading edge speech processing technique to automatically extract and perceptually classify environmental noises towards those new standards.