Generating one phone from one TIMIT speaker

I’ve been trying to repeat my experiment from here but with data from one TIMIT speaker. That is, the goal is to train only on the data of one speaker going ‘aa’ (the middle third of every such phone, to try to reduce coarticulation issues) and get a network that generates an ‘aaa’ sound.

Here is the best generated sound: wav yaml

The network predicts the next three samples form the previous 160, has two rectified linear hidden layers and L2 regularization with parameter 10^(-4). It was trained with vanilla SGD and learning rate 0.3, batchsize 100. The input data was from speaker number 104, and was normalized by the maximum of the phone where each example came from. For the synthesis, the next sample was generated according to normal distribution with mean given by network output and standard deviation given by empirical MRSEs multiplied by 0.1. (It’s not mathematically right, but adding this multiplier makes the generated sound better. With a multiplier of 1 I get static.)

The above wav sounds kind of OK, but it’s not robust at all. Here is the same model generating sound when trained on data for speaker 179. It sounds terrible.

The MSEs of the two models (with data renormalized by empiricial standard deviation of all of TIMIT) are:

speaker number train mse valid mse
104 0.05 0.016
179 0.15 0.08

This suggests I have an overfitting problem (the gap between training mse and valid mse is much larger than Jean Philippe’s). In retrospect this is not very surprising since I’m training on very little data.

It also suggests that the data from speaker 179 is somehow much harder to “imitate” than the data from speaker 104. Not sure why that would be the case.

The next thing I’ll do is see if I can reduce overfitting by using stronger regularization.

Advertisements
Standard

2 thoughts on “Generating one phone from one TIMIT speaker

  1. Pingback: Second MLP architecture for next sample prediction – Part II | IFT6266 Project on Representation Learning

  2. Pingback: Second Experiment — Say ‘aa’ vanilla MLP | IFT6266

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s