Sidney Gharib - Software Developer
I just graduated from Polytechnique Montréal with a bachelor in Software Engineering. I am a passionate software developer and researcher with expertise in full-stack development, machine learning, and computer vision. I enjoy building applications that solve real-world problems and contributing to the scientific community through research and publications.Scientific Publications
My contributions to research and academia
MuViH: Multi-View Hand gesture dataset and recognition pipeline for human–robot interaction in a collaborative robotic finishing platform
Robotics and Computer-Integrated Manufacturing • 2025
RGB-D Cameras
In my research, I had the opportunity to work with Intel RealSense RGB-D cameras, which allowed me to explore 3D depth perception capabilities. These cameras feature dual sensors that enable distance measurement for each pixel, effectively adding depth information to create three-dimensional images.
Calibration
Following the limitations discovered with RGB-D depth sensing, I transitioned to a stereo camera approach as recommended by my supervisor. This method doesn't rely on the built-in depth sensor of the RealSense RGB-D cameras but instead uses a combination of two cameras to create the 3D effect through stereoscopic vision. I calibrated the stereo system using two RGB-D cameras (though without utilizing their depth sensing capabilities) by determining the rotation and translation matrices needed to transform coordinates from the primary camera to the secondary camera
Finger detection
Once the cameras were properly calibrated, I utilized Google's MediaPipe library to detect my index finger within the video stream. MediaPipe provides robust hand landmark detection, identifying 21 key points on each detected hand, including three specific points on the index finger that were crucial for my application. The system works by identifying these key points on both camera feeds simultaneously.
Finger direction in 3D space
After detecting the index finger key points in both camera feeds, I implemented a program to reconstruct 2D points into 3D space using stereo calibration data. This process was applied to all three index finger points detected by MediaPipe.
Real life application
The complete system was ultimately tested in a collaborative robotics (cobotics) room designed to facilitate construction tasks for operators. The project represents a significant step toward improving factory productivity through AI-powered gesture recognition
Work Experience
My professional journey
Featured Projects
Some of my notable personal and professional project
Awards
Research awards
Walphyre - Software Company
Walphyre is a software company cofounded by me and my brother Jason Gharib