Machine Learning & Image Processing using simulated microscope images and CNNs
Simulated electron microscope imagery and built rudimentary convolutional neural network models to automate processing and analysis tasks. Used these models as a foundation for labeling-oriented computer vision workflows where image features would support classifying or interpreting visual content. Iteratively tested and validated the resulting outputs for consistency with the underlying image simulations. • Generated/simulated electron microscope images for ML workflow development • Designed CNN architectures for image processing and analysis • Evaluated model outputs for accuracy and consistency • Developed practical pipelines for converting scientific imagery into ML-ready inputs