Zeineb NEJIM

- Developed computer vision solutions for video applications in non professional sports: cameraman effect, Padel stats (CNN, PyTorch, TensorFlow). - Continuously tested and validated models, algorithms and deliverables. - Deployed the developed solutions into production (API, Jenkins, docker). - Collaborated with cross-functional teams to ideate and prototype innovative video solutions for non-professional sports challenges. - Developed and maintained documentation for computer vision projects and processes (Github).

3D microscopy image analysis: from image enhancement to quantitative analysis of arterial tissues Tasks: - Conducted research on aneurysms microstructure and second harmonic generated images. - Enhanced microscopic 3D images of both healthy and aneurysmal aortic wall (python, mathematical morphology). - Estimated collagen fibers orientations from 3D microscopic images using deep learning as an innovative approach (CNN, Tensorflow). - Created a synthetic dataset to train the developed model (python).

Spatial relations determination and actions prediction using deep learning Tasks: - Determined the spatial relations between the driver and surrounding cars (python). - Conducted research on trajectories estimation using deep learning. - Estimated cars trajectories (LSTM, keras).

Joined the Quality Assurance team working on a software for Electronic Document Management.

Analysis and recognition of human gestures using 3D geometric representation Tasks: - Used the Kinect to extract trajectories of the different joints composing the skeleton (C#). - Created a dataset of 8 motions (C#). - Classified the carried out motions among the created dataset (C#).

Designing a cranking simulator of a steam turbine ALSTOM 240 MW using WinCC flexible 2008 and Step7 to implement the proposed program on a S7-300 PLC

Implementing a PID controller on a real-time PXI system from National Instruments in order to detect terahertz pulse amplitude and position in Real-Time