Simone Bonato

• Engineered and implemented custom semantic and instance segmentation models along with data pipelines for enhanced caries detection in dental X-ray imaging.

• Developing advanced instance segmentation models for precise room detection in architectural floor plans.

• Designing real-time fall detection systems that integrate sensor and camera data to improve safety monitoring.

• Executed diverse data analysis initiatives to derive actionable insights and support evidence-based decision-making.

• Supervising a vessel segmentation project utilizing OCT data to augment diagnostic precision.

• Graphons as efficient Graph Embedding (KTH - SEB)
• Coded various types of Image Processing algorithms (Feature Matching and Detection, Recognition with Vocabulary Tree, Visual Recognition using CNN)
• OpenCv for Image Processing
• Colorful Image Colorization (project): trained a CNN from scratch using Keras to reproduce the results obtained in the homonym paper
• Coded a CNN and a RNN from scratch
• Deep Learning in Data Science (KTH)

• Working on my degree project on Medical Image Segmentation of the Optic Nerve Sheath from head CT scans
• Using nnU-Net to perform segmentation of Optic Nerve Sheath on head CT scans
• Extracting measures from ONS (diameter, volume, etc.) and correlation analysis with Glasgow Outcome Scale after 6 months
• Comparison with other segmentation methods for medical images (Unetr, DeepLabV3, V-Net)