Raman spectra contain abundant information from molecules but are difficult to analyze, especially for the mixtures. Deep-Learning-Based Components Identification for Raman Spectroscopy (DeepCID) has been proposed for the problem of components identification. Convolution Neural Network (CNN) models have been established to predict the presence of the components in the mixtures.
Python 3.6 is recommended.
1.Numpy
pip install numpy
2.Scipy
pip install Scipy
3.Matplotlib
pip install Matplotlib
git clone at：https://github.com/xiaqiong/DeepCID.git
Since the model exceeded the limit, we have uploaded all the models and the information of mixtures to the Baidu SkyDrive and Google driver.
Download at: Baidu SkyDrive or Google driver
1.Training your model
Run the file 'one-component-model.py'.The corresponding example data have been uploaded to the folder named 'augmented data'.
2.Predict mixture spectra data
Run the file 'DeepCID.py'.An example mixture data have been uploaded at Baidu SkyDrive (named 'mixture.npy', 'label.npy' and 'namedata.csv').Download the model and these example data，DeepCID can be reload and predict easily.
Zhi-Min Zhang: zmzhang@csu.edu.cn