Soft Processing of Audio Signals
Andrzej Czyżewski · Studies in fuzziness and soft computing · 1998
The aim of the presented research is to develop and to test some digital signal processing systems applicable to modern telecommunications. A special feature of the elaborated systems is the improvement in performance of audio signal processing algorithms obtained through the use of some soft computing methods based on rough sets, fuzzy logic and neural networks. The engineered and tested digital signal processing systems enabled some comparative studies of the effectiveness of algorithms based on soft computing. The results of speaker-independent recognition of digits and of noise removal from speech and music signals will be presented. Some general conclusions concerning the application of intelligent decision systems to real-time signal processing will be added. These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.