Automatic speaker localization based on speaker identification -A smart room application

This paper presents a system of speaker localization, based on the speaker recognition task. The main application could be the supervision of smart hospital-rooms, where the different crying patients (speaking patients) are localized thanks to their voice features inside the hospital. The proposed method uses the information given by the microphones, placed in different positions, in order to determine the position of the active speaker (person/patient who is crying) and try to supervise the audio-visual recording (eg. supervision of smart-rooms). The speaker identification approach is based on Second Order Statistical Measures (SOSM), which employs the vocal characteristics of the speaker (Mel Frequency Spectral Coefficients). Experiments are conducted on several scenarios containing one, two or three speakers in the smart-room. Results show that the proposed approach is interesting and easy to implement.