1pSP4. On the problems in applying Bell's blind separation to real environments.

Session: Monday Afternoon, December 2

Time: 2:45


Author: Tomohiko Taniguchi
Location: Dept. of Elec. Eng., Nagoya Univ., Naogya-shi, 464-01 Japan
Author: Hani Yehia
Location: Dept. of Elec. Eng., Nagoya Univ., Naogya-shi, 464-01 Japan
Author: Kazuya Takeda
Location: Dept. of Elec. Eng., Nagoya Univ., Naogya-shi, 464-01 Japan
Author: Fumitada Itakura
Location: Dept. of Elec. Eng., Nagoya Univ., Naogya-shi, 464-01 Japan

Abstract:

A blind separation method proposed by Bell et al. [Proc. ICASSP 95 (1995)] is a novel approach for source separation based only upon the assumption that the source signals are statistically independent of each other. In this paper, only simple computer-simulation results were described. The conditions of the experiments were extended and the method was evaluated under real conditions. In the first experiment, the SNRs of source signals are controlled by adding white Gaussian noise in order to clarify to what extent the source signals can be regarded as independent. The results clarify that SDR (i.e., signal-to-deviation ratio between the source and the separated signals) is proportional to the SNR of source signals when the SNR is below 20 dB, and the SDR saturates at 40 dB when the SNR is better than 20 dB. In the second experiment, the lengths of the signals are controlled to find the sufficient source signal length of processing. From the results it is found that the source signal length must be more than 2000 points to achieve an SDR of 30 dB.


ASA 132nd meeting - Hawaii, December 1996