Camera Doodler
Touchless Creativity: where your ideas flow without ever touching a screen!
Created on 21st March 2024
•
Camera Doodler
Touchless Creativity: where your ideas flow without ever touching a screen!
The problem Camera Doodler solves
Camera Doodler
Camera Doodler is a groundbreaking webcam-based system designed to revolutionize online teaching and collaboration, particularly on platforms such as Zoom or Google Meet. By eliminating the need for physical writing pads, it offers a range of innovative capabilities:
Features:
Real-time Drawing: Users can draw and annotate directly on-screen using hand gestures captured by the webcam. This feature enables dynamic explanations and visual aids during online sessions.
Mathematical Expression Evaluation: The system allows users to input mathematical expressions through gestures, enhancing the interactive learning experience. Students can solve problems and visualize mathematical concepts effortlessly.
Target Audience:
Students: In interview tests or virtual classrooms, students often face challenges in explaining concepts due to the lack of writing tools. Camera Doodler, when integrated as a plugin with conference apps like Zoom or Google Meet, empowers students to articulate their ideas effectively through gesture-based drawing and expression evaluation.
Teachers: Educators can utilize the system to visually explain complex concepts, saving time and resources. They can easily draw diagrams, write equations, and annotate content, enriching their online teaching sessions. Moreover, the ability to save notes and export them as PDFs with a single click streamlines the process, making teaching more efficient and cost-effective.
Benefits:
Camera Doodler enhances the online learning environment and promotes accessibility and inclusivity by providing a seamless way for users to interact and communicate visually, regardless of their physical location or access to traditional writing tools.
Challenges we ran into
Overcoming Challenges in Building the Project
During the development of our webcam-based hand movement tracking system, we encountered specific challenges and hurdles, along with our solutions:
Augmenting Dataset for Handwritten Digits and Mathematical Symbols:
Challenge: Although we found a dataset with a large number of images, augmenting it to increase diversity while ensuring distinctiveness for training was tricky. Simply altering positions, rotations, and scales wasn't generating sufficiently unique images.
Solution: To overcome this, we implemented more advanced augmentation techniques such as elastic distortions, random affine transformations, and adding noise to the images. This approach significantly enhanced the diversity of the dataset while ensuring each image remained distinct for effective training.
Encrypting Video Feed for Safety:
Challenge: Ensuring the safety and privacy of users' video feeds was paramount. However, encrypting the video feed without hindering the system's performance posed a challenge.
Solution: We addressed this challenge by implementing a functionality that prompts users to grant permission before accessing the camera feed. This not only ensures user consent but also indirectly enhances safety by preventing unauthorized access to the camera. Additionally, we integrated encryption protocols to safeguard the video feed during transmission, thereby ensuring end-to-end security.
Tracks Applied (1)
AI and ML
Technologies used