Inspiration
My Big sister, Delta, was born with a hearing impairment. I watched her and others like her struggle every day—education became a problem as they couldn't understand, words slipping through fingers like sand. They tried so hard to understand, but the world spoke too fast. One night i thought: What if words weren't just sounds? What if she could see them? That's how the 3D Alphabet Visualizer was born—a tool to turn speech into something tangible, allowing impaired individuals to communicate without being left out and in education to bring light and hope to individuals feeling left out.
What it does
The Visualizer listens to spoken words and transforms them 3D letters. Say "hello," and a 3d model signing "H-E-L-L-O" appear, twisting gently in the air. Features include:
- Real-time speech-to-3D-letter conversion
- Interactive letter manipulation
- Signing avatar companion
- Multi-language support in development
How we built it
We started from scratch—two beginners with a dream: Voice to Vision: Using the Web Speech API, we captured spoken words and filtered them into clean letters.3D Letters, Zero Experience: We spent nights learning Blender, sculpting each letter model by hand (our first "A" looked like a melted ladder).The Companion: We added a friendly animated guide—a signing avatar—to help bridge the gap between letters and meaning.
Challenges we ran into
3D Animation Nightmares: We didn't know a vertex from a voxel. Our early models glitched, flipped, or vanished mid-air.Speech Errors: Background noise turned "cat" into "zzat," making my sister giggle but us groan.Sign Integration: We wanted signing to work backwards—capturing hand movements and turning them into speech. The first test? My flailing arms registered as "banana."
Accomplishments that we're proud of
From Zero to 3D Heroes: We went from struggling with basic shapes to animating smooth, interactive models.
What we learned
*Imperfect > Perfect: * Our janky first models taught us more than any tutorial. *Accessibility is Magic: * A small tweak—like slowing letter display—made all the difference . *Sign Language is a Universe: * We barely scratched the surface of ASL, let alone other global signing systems.
What's next for 3D model asl sign
Beyond ASL:
Adding support for:
- [ ] BSL (British Sign Language)
- [ ] LSF (Langue des Signes Française)
- [ ] More global sign languages
Because every sign language deserves a voice.
Camera That Listens to Hands:
Developing a webcam mode that:
- Watches signing movements
- Translates signs in real-time
- Speaks the words aloud
Helping hearing people understand her world. Community Voices:
- Open-sourcing the project
- Crowdsourcing language packs
- Building a model repository
So others can add:
- Their languages
- Their letters
- Their sister's smile
"My sister signs to me now, her hands dancing: 'More letters, please.'
And we're just getting started."
Log in or sign up for Devpost to join the conversation.