This project aims to create a software program, based on a Convolutional Neural Network, which can distinguish male from female voices based on a short audio recording of a human voice. Upon receiving an audio input, the software will return a classification 'male' or 'female'.
The software has been trained using a large number of recordings of human adults saying a single digit i.e. 0-9.
The data used to train and validate this solution is AudioMNIST. It is widely used to train deep and/or convolutional neural networks for speech-recognition tasks. In this project the dataset is used to train a CNN to classify audio input as either male or female depending on the speaker.
The dataset was created by Becker et al and is available here: https://github.com/soerenab/AudioMNIST. The creators of the dataset request that projects using their work cite the following paper:
| Paper Title | Interpreting and Explaining Deep Neural Networks for Classification of Audio Signals |
|---|---|
| Journal | CoRR |
| Volume | abs/1807.03418 |
| Year | 2018 |
| ArchivePrefix | arXiv |
| Eprint | 1807.03418 |
| Authors | Becker, Sören and Ackermann, Marcel and Lapuschkin, Sebastian and Müller, Klaus-Robert and Samek, Wojciech |
These data must be preprocessed in order to make them suitable for processing by a convolutional neural network.
Preprocessing includes adding a specific form of emphasis, as well as padding. Padding ensures that the resulting input
has the same length irrespective of audio sample.
The preprocessing code in included in this project in the Jupyter notebook
preprocess_audiomnist.ipynb.
Test audio was obtained from the author's work environment. These audio samples are comparable to the AudioNIST samples, in that they are approximately 1s in length and are recordings of adults verbalising the digits 0 through 9. One difference is that speakers provided samples in different languages. This was a quick way to obtain a larger test set. If the model is working well, and as long as speakers do not fundamentally change their voices when speaking the foreign language, these samples should provide a good test of the model's capability.
The model chosen to perform this task is a Convolutional Neural Network (CNN). The CNN does not process audio samples (in the wav format) directly, but instead processes an image representation (in the jpg format) of the original audio. The image represents a so-called Mel-frequency cepstrum (MFC). For example, the following MFC image represents a female saying the digit "zero":
These RGB images have shape 98x12x3, and the CNN is specifically designed to process shapes of this kind.
I performed an exhaustive grid search of the following hyperparameters and values:
| Parameter | Value 1 | Value 2 | Value 3 |
|---|---|---|---|
| Epochs | 5 | 10 | 20 |
| Learning rate | 0.1 | 0.01 | 0.001 |
| Gamma | 0.1 | 0.3 | 0.7 |
Gamma is the multiplicative factor of learning rate decay. The learning rate was adjusted every 5 epochs.
TODO
A summary of your results and what you can learn from your model
The project tries to exploit specialised hardware, if present. Training was performed on a Mac M1 Max with both CPU and MPS (Mac M1 GPU). Switching from CPU to GPU results in a 3-fold increase in training speed. The project will automatically detect cuda or mps hardware if present. Program arguments can be used to disable these options of required.
Mike Mannion B.Sc. (hons) MBA [email protected]