BRIDGE-AI: Sign Language Interpreter
Artificial intelligence has the power to transform lives, especially for people with disabilities. But AI models are only as good as the data they are trained on. After spending six months developing a prototype for assistive technology, I realized that to make it truly useful, I needed a larger, more accurate dataset. The best way to achieve this? Engaging with the community that would actually use it.
To create a more inclusive AI model, I collaborated with the Persons with Disabilities (PWD) community in Naga City. With the support of the local government unit (LGU), I organized a workshop for the deaf and hearing-impaired community, introducing them to recent advancements in assistive technology. More importantly, I shared my prototype—an AI-powered system designed to recognize sign language gestures.
This workshop wasn’t just about presenting technology; it was about fostering engagement. The attendees shared valuable insights about their daily experiences, the challenges they face, and the role technology could play in their lives. These conversations underscored the importance of developing AI tools that are not just accurate but also culturally and contextually relevant.
With the backing of city officials, I secured permission to collect datasets of different sign language gestures. However, I quickly realized that traditional data collection methods—such as requiring participants to visit a research lab—would be impractical. Many people wouldn’t have the time or resources to participate under these constraints.
To solve this, I developed a mobile application using Flutter that allowed participants to record sign language gestures using their own phones. This made the data collection process more accessible, flexible, and scalable, as participants could contribute at their own pace and in familiar environments. By leveraging Flutter, I ensured that the app worked seamlessly across both Android and iOS devices, making it easy for a diverse group of users to participate.
Beyond data collection, this experience was profoundly personal. I had the opportunity to connect with incredible individuals, learn about their culture, and gain a deeper appreciation of the deaf and hearing-impaired community. Their perspectives challenged me to think beyond just technological solutions and focus on how AI can integrate into and enhance real lives.
The collaboration reinforced a crucial lesson: AI development shouldn’t be a one-way street. The most effective and ethical AI models are built in partnership with the communities they aim to serve.
This initiative is just the beginning. The next steps involve refining the model, testing it with real users, and exploring potential applications in education, accessibility, and communication. As AI continues to evolve, community-driven data collection will be key to making technology more inclusive and impactful.
By working together, we can ensure that AI serves everyone—especially those who stand to benefit the most.