Unofficial implement with paper SpeakerGAN: Speaker identification with conditional generative adversarial network

Introduction

This repository is about paper SpeakerGAN , and is unofficially implemented by Mingming Huang ([email protected]), Tiezheng Wang ([email protected]) and thanks for advice from TongFeng.

SpeakerGAN paper

SpeakerGAN: Speaker identification with conditional generative adversarial network， by Liyang Chen , Yifeng Liu , Wendong Xiao , Yingxue Wang ,Haiyong Xie.

Usage

For train / test / generate:

python speakergan.py

You may need to change the path of wav vad preprocessed files.

Our results

acc: 94.27% with random sampled testset. 

acc: 93.21% with fixed start sampled testset.

using model file: model/49_D.pkl

acc: 98.44% on training classification accuracy with real samples.

There is about 4% gap on testset lower compared to paper result. We can't find out the reason. We want your help !

Details of paper

The following are details about this paper.

================ input ==================

1. feature: fbank, 8000hz, 25ms frame, 10ms overlap. shape:(160,64)

2. dataset: librispeech-100 train-clean-100 POI:251

3. data preprocess: vad、mean and variance normalization, shuffled.

4. 60% train. 40% test.

================ model architecture ==================

1. dataflow: data -> feature extraction -> G & D

2. model architecture:

   G: gated CNN, encoder-decoder, Huber loss + adversarial loss

   D: ResnetBlocks, template average pooling, FC, softmax, crossentropy loss + adversarial loss

3. G: shuffler layer, GLU

4. D: ReLU

================ training ==================

1. lr: 0-9, 0.0005 | 9-49, 0.0002

2. L(d): λ1 λ2 = 1

3. batch_size: 64

4. D_train steps / G_train steps = 4

5. Ladv Loss: Label smoothing, 1 -> 0.7 ~ 1.0, 0 -> 0 ~ 0.3

======== not sure or differences with paper ========

1. weights,bias initialize function, use: xavier_uniform and zeros

2. pytorch huber_loss： + 0.5 to be same with paper. but no implement here.

3. for shorter wav, paper: padded with zero. we: padded with feature again.

4. gated cnn architecture.

5. we use webrtcvad mode(3) for vad preprocess.