Inspiration
Difficulties experienced by individuals with conditions like Parkinson’s disease highlight how physically demanding everyday web browsing can be when interfaces rely on precise mouse and keyboard control. Tasks such as scrolling, clicking, and switching tabs can become slow and exhausting for many users with motor impairments or limited dexterity. These barriers reflect a broader gap in accessible interaction design. VoxSurf was created to address this need a multimodal Chrome extension that enables hands-free web navigation through voice, gesture, and motion.
What it does
VoxSurf is a multimodal Chrome extension that lets users navigate the web without touching a mouse or keyboard. It translates voice commands, hand gestures, and body motion into real browser actions like opening websites, switching tabs, scrolling pages, and searching online. By turning natural human input into seamless browser control, VoxSurf makes web navigation faster, hands-free, and accessible for people with different physical abilities.
How we built it
We built VoxSurf as a React/Vite Chrome extension with a content-script runtime, combining OpenCV gesture and motion tracking with in-browser speech recognition. These inputs are unified into real-time browser actions like tab control, navigation, and scrolling.
Challenges we ran into
YouTube Shorts interaction layers made click targeting inconsistent
Accomplishments that we're proud of
Automated dist sync to Windows path so every build is instantly testable
What we learned
Fewer, clearer gestures beat “feature-rich” gesture sets
What's next for VoxSurf
Telemetry/debug mode to tune missed or false gestures quickly
Built With
- openai
- react
Log in or sign up for Devpost to join the conversation.