A virtual board controlled by hand gestures.

• Problem Statement
• Idea / Solution
• Dependencies / Limitations
• Future Scope
• Questions & Answers
• Setting up a local environment
• Usage
• Technology Stack
• Authors
• Acknowledgments

Teachers need to use multiple screens to display the material and their selves.

Allow the teachers to display all the materials and their selves in a smart way with just one monitor, and without having to use any extra solutions, like a green screen.

Teachers are having a hard time making students more focused and interested during the online lecture.

Teachers should have a tool that allows them to grab the attention of students and make them more focused, without any effort on the teacher part.

• An application that enables the teachers to display both their camera feed and class material simultaneously with an intuitive input design, hands position and gestures, that allows the teacher to focus more on the teaching side and less on the technical side.

• Based on hand recognition, which is usually implemented on hardware optimized parts, to use it on every camera might require a bit of optimization.

• Allow students to interact with teachers via the app
• Add attention mechanisms
• Convert the desktop app to a web app
• implement Brownian Motion to make objects displacement more natural

todo

1. Unity 2019
2. Visual Studio 2019
3. Python 3.8
4. OBS, OBS-VirtualCam

Install required libraries

OpenCV - library of programming functions mainly aimed at real-time computer vision

pip install opencv-python

Mediapipe - the simplest way for researchers and developers to build world-class ML solutions and applications

pip install mediapipe

Scikit-learn - Simple and efficient tools for predictive data analysis in python

pip install -U scikit-learn

1. download OBS - https://obsproject.com/download
2. download OBS Virtual Camera - https://github.com/CatxFish/obs-virtual-cam/releases

• download unity 2019 - https://unity.com/
• everything is pretty much ready out of the box.

1. Setup a virtual camera from your main camera
2. Run Assets/ML Server/main/decodeFromLocal.py
3. Open the project in Unity and run s1.scene

• Unity - World’s best real-time development platform.
• Mediapipe - the simplest way for researchers and developers to build world-class ML solutions and applications.
• Scikit-learn - Simple and efficient tools for predictive data analysis in python.

• @bshara23
• @ahmad-masalha
• @aimanouda2309

Not one yet