This project used regression and segmentation models to obtain the monolayer crystal coverage of WSe2 from their micrographs.
- to explore the behavior of regression models compared to segmentation models for characterizing micrographs, such as determining the thin film crystal growth on a substrate, quantified by crystal coverage
- to investigate how pretraining domains and different modes of transfer learning impact the capabilities and reliability of models at inferencing crystal coverage of samples not included in the training
Processed data for this project can be found at https://doi.org/10.5281/zenodo.10784189 8432222 Raw data can be found at https://m4-2dcc.vmhost.psu.edu/list/data/RVJkDr8j1RPU
- Regression models: notebooks/regression
- Segmentation models: notebooks/segmentation
- Plots: notebooks/plots
- codes: codes
- plots and data from models: Result
- Trained models: Models .
@article{moses2024crystal, title={Crystal growth characterization of WSe2 thin film using machine learning}, author={Moses, Isaiah A and Wu, Chengyin and Reinhart, Wesley F}, journal={Materials Today Advances}, volume={22}, pages={100483}, year={2024}, publisher={Elsevier} }