TensorFlow implementation of VGGish-BiGRU for IEEE Access paper "Lung Sound Recognition Algorithm Based on VGGish-BiGRU", Lukui Shi, Kang Du, Chaozong Zhang, Hongqi Ma and Wenjie Yan.
The respiratory sounds from "ICBHI 2017 Challenge" database.
The parameters of VGGish network provided by Google are directly transferred to the model and frozen.
Please download "vggish_model.h5" from vggish-keras.