📖 Project Story: IndicSign
🧠 Inspiration
The idea for IndicSign was inspired by a visit to a school for mute and hearing-impaired individuals in Delhi. Meeting these differently-abled heroes and witnessing their resilience and brilliance was both humbling and eye-opening. It highlighted how communication remains a significant barrier, and how technology could be a powerful enabler to bridge that gap. This experience motivated me to work on a college project that could genuinely make a difference.
🔧 How I Built It
IndicSign was developed as a college project using:
- Flask for the backend framework
- HTML/CSS for the front-end interface
- Google’s Gemini API for multimodal gesture recognition
- Computer vision to process hand gesture inputs
Originally, I aimed to train and host a large ML model locally. However, due to hardware limitations and deployment challenges, I opted to integrate the Gemini API, which offered a scalable and intelligent solution without the need for local model hosting.
🚧 Challenges Faced
- Hosting Constraints: Deploying a heavy model on limited infrastructure was not feasible, leading to the switch to a cloud-based API.
- Real-time Processing: Maintaining responsiveness and low latency for gesture recognition was essential for usability.
- Inclusive Design: Ensuring the interface was clean, accessible, and intuitive for all users was a key focus.
💡 What I Learned
- Seamless integration of AI APIs (like Gemini) into web applications
- Building inclusive tech solutions with empathy and user needs in mind
- Managing backend/frontend development in sync using Flask and HTML/CSS
- The importance of designing for real-world impact, especially for underserved communities
🌈 Final Thoughts
IndicSign is more than just a tech solution—it's a tribute to the strength and spirit of the differently-abled community. Through this project, I’ve learned that even small innovations can open big doors when built with purpose and compassion.