Signify

• 

• 

  Part of the code from the AI model and the sign-language translator

Inspiration

Many people nowadays suffer from various disabilities, making them unable to feel or to express themselves as we do. Because of this they often feel isolation and discrimination.We wanted to create a seamless way to bridge this gap using AI and computer vision.

What it does

Signify translates sign language into text in real time, allowing users to communicate quickly and effortlessly. By analyzing hand movements, it converts gestures into readable text instantly. It also has a face recognition feature, allowing people with sight problems to recognize people they know.

How we built it

We developed a recursive neural network from scratch using Tensorflow to process motion data. MediaPipe and OpenCV handle real-time hand tracking, while Pygame powers the intuitive UI. The Face Recognition feature is implemented using a deep neural network pre-trained on the VGG face dataset, and uses OpenCV to get the visual features.

Challenges we ran into

Building a recursive neural network with no prior experience in AI was a challenge, requiring extensive tuning for accuracy. Real-time motion tracking also needed optimization to ensure smooth and responsive translation.

Accomplishments that we're proud of

We successfully implemented an AI-powered sign-to-text system with no external ML libraries. Creating a user-friendly experience while maintaining accuracy and speed was a major achievement.

What we learned

We gained deep insights into neural networks, computer vision, and real-time processing. Additionally, we learned how to balance efficiency, accuracy, and usability in AI-driven applications.

What's next for Signify

We plan to expand Signify with multi-language support, voice synthesis, and mobile integration. The final goal would be to connect the software with a camera embedded on accessories like glasses, to improve accessibility and inclusivity while maintaining a light and efficient usage . Further improvements in accuracy and real-time performance will make it even more powerful and accessible.

Built With

• opencv
• pygame
• python
• tensorflow