Blind signal separation for convolved nonstationary signals
โ Scribed by Mitsuru Kawamoto; A. K. Barros; A. Mansour; Kiyotoshi Matsuoka; Noboru Ohnishi
- Publisher
- John Wiley and Sons
- Year
- 2000
- Tongue
- English
- Weight
- 333 KB
- Volume
- 84
- Category
- Article
- ISSN
- 1042-0967
No coin nor oath required. For personal study only.
๐ SIMILAR VOLUMES
## Abstract An adaptive blind source separation algorithm for the separation of convolutive mixtures of cyclostationary signals is proposed. The algorithm is derived by applying natural gradient iterative learning to a novel cost function which is defined according to the wide sense cyclostationari
The development of an acoustic traffic signal for blind pedestrians is described. The WAL K and WAITsignals used represent a compromise between many contradictory demands, e.g. easy localisation and detection in a background of traffic noise, minimum annoyance, maximum attenuation into dwellings and
In this paper we propose a new two-dimensional least mean squares algorithm (2D-LMS) which is able to track nonstationarities in both vertical and horizontal directions with a computational load comparable to 1D-LMS methods of the same number of weights. The main difference of our method consists in