README
¶
speechrecog
This is a set of tools for implementing speech recognition. This is the first time I have played with speech recognition, so I am not exactly sure what will be needed. Nonetheless, here is what I have so far:
- An MFCC package
- A web app for recording and labeling speech samples
- CTC recurrent neural net training
License
This is under a BSD 2-clause license. See LICENSE.
Directories
¶
| Path | Synopsis |
|---|---|
|
Package ctc implements Connectionist Temporal Classification for training models (typically neural networks) to predict output sequences.
|
Package ctc implements Connectionist Temporal Classification for training models (typically neural networks) to predict output sequences. |
|
Package mfcc can compute Mel-frequency cepstrum coefficients from raw sample data.
|
Package mfcc can compute Mel-frequency cepstrum coefficients from raw sample data. |
|
Command mfcc-graph produces HTML graphs of MFCC coefficients for an audio file.
|
Command mfcc-graph produces HTML graphs of MFCC coefficients for an audio file. |
|
Command recorder is a web app for recording speech clips.
|
Command recorder is a web app for recording speech clips. |
|
Package speechdata facilitates loading and saving databases of labeled speech samples.
|
Package speechdata facilitates loading and saving databases of labeled speech samples. |
Click to show internal directories.
Click to hide internal directories.