There are some codes for the paper DeepMIR: a versatile and accurate component identification method for high-throughput mid-infrared spectra. We developed a method named Deep learning-based library search method for Mid-infrared spectroscopy (DeepMIR). Other methods for comparison in the paper are also presented. Organic solvents datasets are public for example applications.
Before running codes on your own computer, make sure you have established an environment required by this project.
conda env create -f requirements.yml -n myenv
activate myenv
You can see the example.ipynb from the "src" folder for the utilization of DeepMIR. If you want to try the codes on your own computer, please download the whole project and the Organic Solvents dataset from the Releases. Try other sub-datasets based on the example.ipynb.
A web server of DeepMIR is established. Please refer to the paper for more details.
LinTan lintan_2000@csu.edu.cn