Low-power implementation of an HMM-based sound environment classification algorithm for hearing aid application

Automatic program switching is a future trend for digital hearing aids. To realize this function, a solution for sound environment classification is required. This paper presents an HMM-based sound environment classifier that is implemented on a low-power DSP system designed for hearing aid applications. Our experimental results show that it is capable of distinguishing four sound sources (i.e. speech, music, car noise, and babble) with more than 95% accuracy rate and consumes only 0.225 mW of power.

[1]  Henning Puder,et al.  Signal Processing in High-End Hearing Aids: State of the Art, Challenges, and Future Trends , 2005, EURASIP J. Adv. Signal Process..

[2]  Arne Leijon,et al.  An efficient robust sound classification algorithm for hearing aids. , 2004, The Journal of the Acoustical Society of America.

[3]  J M Kates,et al.  Classification of background noises for hearing-aid applications. , 1995, The Journal of the Acoustical Society of America.

[4]  David V. Anderson,et al.  Audio classification and scene recognition and for hearing aids , 2005, 2005 IEEE International Symposium on Circuits and Systems.

[5]  Hamid Sheikhzadeh,et al.  An ultra low power, ultra miniature voice command system based on Hidden Markov Models , 2002, 2002 IEEE International Conference on Acoustics, Speech, and Signal Processing.

[6]  Norbert Dillier,et al.  Sound Classification in Hearing Aids Inspired by Auditory Scene Analysis , 2005, EURASIP J. Adv. Signal Process..

[7]  Lawrence R. Rabiner,et al.  A tutorial on hidden Markov models and selected applications in speech recognition , 1989, Proc. IEEE.