https://github.com/billryan/toneclassifier
Tone Classifier for Machine Learning course at SJTU
https://github.com/billryan/toneclassifier
Last synced: 9 months ago
JSON representation
Tone Classifier for Machine Learning course at SJTU
- Host: GitHub
- URL: https://github.com/billryan/toneclassifier
- Owner: billryan
- Created: 2014-06-14T01:55:18.000Z (over 11 years ago)
- Default Branch: master
- Last Pushed: 2014-06-21T02:21:18.000Z (over 11 years ago)
- Last Synced: 2025-03-25T19:01:48.307Z (10 months ago)
- Language: Python
- Homepage:
- Size: 17.8 MB
- Stars: 6
- Watchers: 3
- Forks: 0
- Open Issues: 1
-
Metadata Files:
- Readme: README
Awesome Lists containing this project
README
This dir contains the data for training/testing a tone classifier.
train/ training data
test/ test data
Each dir includes the f0 and engery features extracted from
corresponding wav files. Each wav file is the pronunciation of
a single Chinese character.
Note that the length of different wav file may be different.
Hence, the f0/engy files corresponding to different character
may also have different length.
But for the same character, the f0 file will have exactly the
same length as the engy file.
The naming convention for subdir and file names are as below:
1. subdir name indicates the correct tone.
2. filename pattern is
{pron}{tone_lable}.f0
{pron}{tone_lable}.engy
where pron is the pronunciation of the character and tone_lable is the tone of that character, i.e. 1, 2, 3 or 4.
Please design a tone classifer with the f0/engy file as the input and output the tone label.