作者: E. Dellandrea , P. Makris , N. Vincent
DOI: 10.1109/ISPA.2003.1296883
关键词:
摘要: We present in this paper an audio signal classification method based on Zipf and inverse laws. These laws are powerful analysis tools allowing the extraction of information not available by way standard methods. The adaptation Inverse to signals requires a coding these into literary texts, considered as sequences patterns. Because codings first importance since they have bring fore relevant signals, three types been developed, depending representation it is on: temporal, frequential time-scale representations. Once coded, features linked computed. Finally, step aims at identification signals. Four methods well fusion used combine classifiers. In order evaluate our method, we analysed medical corresponding swallowing containing xiphoidal sounds. problem characterize them according gastro-oesophageal reflux pathological state.