This code implements the following paper. Sang-Ick Kang and Sangmin Lee, Improvement of Speech/Music Classification for 3GPP EVS Based on LSTM, Symmetry 2018. https://www.mdpi.com/2073-8994/10/11/605
Using LSTM and EVS's Feature for speech-music classification. Inputs are 12 feartures that extrated from 3gpp evs-codec (Enhanced Voice Service)
I think my code has some error. So I will update code by pytorch maybe until 2019 gone.