PORTFOLIO
In this section, I will present projects I have completed during my studies, thesis work, and professional experience at Rafael, AI Technion Lab or at the SAMPL Lab
Pleural Diseases Assessment and Fatty liver Estimation Using
Sub-Sampled Ultrasound Channel Data (Thesis)
This thesis project focuses on studying prediction methods for pleural effusion and fatty liver estimation using ultrasound channel data. The project is divided into three parts: improving ultrasound images for specific tasks through image generation, and direct prediction methods that bypass image processing.
​
Field : Audio Signal Processing - Computer Vision
Sign to Language Translation
This project aims to translate videos of people using sign language into text. Using the How2Sign dataset, which includes 80 hours of sign language videos and their translations, the goal is to achieve a high BLEU score. An enhanced transformer model is developed, incorporating various preprocessing and post-processing techniques.
​
Field : Computer Vision - NLP - LLM
Cognitive Robotic with Reinforcement Learning
The project aims to optimize a restaurant's operations by training robots to collaborate effectively, work together to improve customer satisfaction, maximize income, and manage restaurant tasks. Using multi-agent reinforcement learning, the robots learn to maximizing both revenue and customer satisfaction in a dynamic environment.
​
Field : Reinforcement Learning
Prediction of Multidisciplinary Meeting Outcomes in Hospital
This project involves predicting the outcomes of multidisciplinary meetings at a hospital specializing in interventional radiology based on a patient's medical record. It was carried out as part of a partnership with the Georges Pompidou Hospital in Paris and ENS Paris-Saclay.
​
​
Field : NLP - LLM
Hospital Ressources Optimization
This project focuses on optimizing the resources of the interventional radiology department at Rambam Hospital in Haifa. The aim is to create a real-time monitoring system that can provide instructions to optimize patient care, task allocation among hospital staff, and treatment times, especially under high pressure and staff shortages.
​
Field : Multi-Agent Reinforcement Learning
Automated Mouses Organs Segmentation
This project aims to create an algorithm for segmenting ultrasound images of mouse organs. The goal is to identify various organs in mouse litters, with the number of mice per litter varying, and to classify multiple organs in the images.
​
​
Field : Computer Vision
Tennis Statistics and Data Capture Application
This IoT and machine learning project focuses on developing a sensor that attaches to a tennis racket and connects via Bluetooth to an app. The app captures real-time data, which is then processed by algorithms and AI models to collect statistics on playing style and performance.
​
Field : IOT - Software Engineering - Machine Learning