InterSign Inspiration InterSign was inspired by the need to bridge the communication gap between the hearing and deaf communities. During the Université sans obstacles hackathon, we met individuals with disabilities for the first time. Listening to their daily challenges deeply moved us and motivated us to create a solution that fosters inclusivity and accessibility, ultimately leading to the birth of InterSign.
What it does Sign Language Learning Tool: Provides a plugin where users input text or gestures to receive videos that teach sign language. LLM-Enhanced Responses in Sign Language: Integrates an LLM capable of responding to user queries via sign language avatars, combining natural language understanding with visual communication. How we built it
Backend: Developed using flask and Python, integrating AI inference models for real-time processing. AI Models: Trained on curated datasets, including videos of ASL signs and natural language models, to achieve accurate translation and gesture recognition. Plugins: Implemented tools for learning sign language and LLM responses with sign language outputs.
Challenges we ran into Acquiring and preprocessing labeled datasets for accurate sign language recognition and synthesis. Ensuring real-time processing speed without compromising accuracy or quality of the avatar animations.
Accomplishments that we're proud of Successfully creating AI-powered avatars that translate spoken language into sign language with natural, fluid motions. Launching a plugin that helps users learn sign language interactively through customized videos. Integrating sign language detection and synthesis into video conferencing platforms, making virtual communication more inclusive. What we learned The critical importance of understanding the lived experiences of our target users to create impactful solutions.
Best practices for building cross-platform applications that balance functionality with accessibility. What's next for InterSign Expand Language Support: Introduce additional sign languages beyond ASL, such as BSL and LSF. Enhanced Avatar Customization: Allow users to personalize avatars for greater engagement. Improved AI Models: Train models on larger datasets to enhance accuracy and fluency in translation. Education Modules: Develop interactive lessons and quizzes to make learning sign language more engaging. Enterprise Solutions: Integrate InterSign into enterprise communication tools like Microsoft Teams and Slack. Community Feedback Loop: Continuously refine the platform based on user feedback to ensure it meets diverse needs.
Log in or sign up for Devpost to join the conversation.