@Author: Su, Ming Yi
@Date: 05/21/2019
@Library: OpenCV, Tensorflow, OpenGL
I applied cutting-edge Machine Learning technique: Convolutional Neural Networks to classify music symbols.
Designed CNN architecture and trained CNN model by tensorflow.
Simulated the training data including illumination, rotation, scale, translation by OpenGL Phong shading.
Increased diversity of training data and implemented image segmentation with music knowledge by OpenCV.
This project is my master project, there are 6 directories.
- music_recognization
- music_simulation_3D
- music_simulation_preprocess
- music_simulation_software
- music_training
- Ouput
Here is the pipeline for my project.