作者: Adam B. Fineberg
DOI:
关键词: Pattern recognition 、 Audio signal 、 Artificial intelligence 、 Feature vector 、 Comparator 、 Centralizer and normalizer 、 Radio spectrum 、 Speech recognition 、 Computer science 、 Normalization (statistics)
摘要: A sound recognizer uses a feature value normalization process to substantially increase the accuracy of recognizing acoustic signals in noise. The includes vector device (110) which determines number values for analysis frames, min/max (120) minimum and maximum each frequency bands, normalizer (130) normalizes with resulting normalized vectors, comparator (140) compares vectors template identify one that most resembles vectors.