https://github.com/parham1998/isolated-digits-recognition
Implementation of Persian Isolated-Digits Recognition with Matlab
https://github.com/parham1998/isolated-digits-recognition
digits-recognition isolated-digits matlab mfcc-features speech-recognition
Last synced: 2 days ago
JSON representation
Implementation of Persian Isolated-Digits Recognition with Matlab
- Host: GitHub
- URL: https://github.com/parham1998/isolated-digits-recognition
- Owner: parham1998
- Created: 2021-06-12T19:34:23.000Z (almost 5 years ago)
- Default Branch: main
- Last Pushed: 2021-10-23T11:33:46.000Z (over 4 years ago)
- Last Synced: 2025-12-12T21:34:51.202Z (6 months ago)
- Topics: digits-recognition, isolated-digits, matlab, mfcc-features, speech-recognition
- Language: MATLAB
- Homepage:
- Size: 328 KB
- Stars: 4
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Isolated-Digits-Recognition (warm-up project!)
Implementation of Persian Isolated-Digits Recognition with Matlab
I recorded (0 to 9) digits two times (one for the test and the other one as a reference) in Persian (you can do that in your language!).
Then extracted their features (5 different feature vectors, one for each question(i).m) and tried to recognize them with **cepstral distance** measure. (see image below)
**Actually, cepstral distance is not a good measure for the speech recognition task because it ignores time warping. for more accuracy, you can use neural networks such as RNNs.**
### notations
question1:
12 real cepstrum coefficients
question2:
12 MFCC (Mel-frequency cepstral coefficients)
question3:
12 MFCC + 1 Energy coefficient
question4:
12 MFCC + 1 Energy coefficient and their first derivatives
question5:
12 MFCC + 1 Energy coefficient and their first and second derivatives
### isolated digits or words recognition flowchart
