Speaker
Description
This study investigates the use of deep learning to enhance Raman spectroscopy analysis for two-dimensional (2D) materials, which are valued for their unique structural properties. Traditional methods for analyzing Raman data are time-consuming and rely heavily on manual interpretation, prompting the need for more efficient approaches. We developed a one-dimensional convolutional neural network (1D CNN) to classify Raman spectra of 2D materials and applied generative deep learning models for data augmentation. A generative adversarial network (GAN) was used to create additional data for bilayer graphene with subtle spectral differences, while a denoising diffusion probabilistic model (DDPM) was employed to improve the dataset quality for complex, multi-class 2D materials. These models significantly enhance classification accuracy, demonstrating the feasibility and effectiveness of deep learning in automating and improving the precision of material characterization, even in cases of limited data. This approach holds significant potential for advancing automated analysis and reducing human intervention in materials science.
