Ines Bouhelal

Data Preparation: Ensuring the compatibility of medical imaging data (NIFTI and DICOM format) with data preprocessing scripts. •Segmentation of glioblastoma brain tumors: Utilized and fine-tuned transformer-based model Segment Anything Model for the task, achieving a dice score of 0.741, a sensitivity of 0.994, and a specificity of 0.841.

Data Preparation: Collected, cleansed, preprocessed and documented a fire dataset guaranteeing data accuracy and integrity as a reliable foundation for data-driven decision-making processes. •Object Detection: Employed YOLO (v3, v4, and v5), and Faster-R-CNN for fire detection, reaching a 70% accuracy rate with the best model and deploying it on the company's robot.

Authentication in Flutter: Designed and established client authentication and custom profiles. •Data Management: Stored client requests and demands using a NoSQL Database Firebase eliminating 10% of paperwork for the company. •App Interface: Accomplished a compelling UI/UX design for the app via Figma and seamlessly implemented it.

3.4 GPA