This is a source code for the paper "unsupervised feature learning basded on deep models for environmental audio tagging"
Better audio tagging code is here:
https://github.com/yongxuUSTC/dcase2017_task4_cvssp