Gayatri Narale

• Developing and systematically evaluating a PyTorch-based deep learning solution for visual anomaly detection using real and synthetically generated image data. • Investigating synthetic data generation and augmentation strategies to increase data diversity and improve model robustness and generalization.

Developed a Python/JupyterHub-based data solution for processing, cleaning and analyzing industrial production data. • Analyzed technical and user requirements and evaluated data-driven approaches for industrial process monitoring and decision support. • Evaluated alternative approaches based on accuracy, computational requirements and practical feasibility.