Multichannel Robot Speech Recognition Database: MChRSR

In real human-robot interaction (HRI) scenarios, speech recognition represents a major challenge due to robot noise, background noise and time-varying acoustic channel. This document describes the procedure used to obtain the Multichannel Robot Speech Recognition Database (MChRSR). It is composed of 12 hours of multichannel evaluation data recorded in a real mobile HRI scenario. This database was recorded with a PR2 robot performing different translational and azimuthal movements. Accordingly, 16 evaluation sets were obtained re-recording the clean set of the Aurora-4 database in different movement conditions. 1. Database Recording The experimental setup used in the database recording employs a PR2 robot which is a state-of-the-art mobile manipulation robot. It has a Microsoft Xbox 360 Kinect sensor mounted on the top. We re-record the clean test set from Aurora-4 database [1] in a meeting room considering different relative movements between the speech source and the robot. A TANNOY 501a loudspeaker was used as the audio source. The recording process was performed with the PR2’s Microsoft Kinect sensor which has a four-microphone array. The re-recording was carried out while the robot was performing translational and head rotation movements simultaneously. Before the recording of each robot movement condition, the background noise was measured and the maximum of equivalent sound pressure level (Leq) over ten minutes was 39dBA.