Inspiration
For the past decade or so, computer interaction has barely had any innovation. The traditional mouse and keyboard have ruled the market. We wanted to innovate on that, so we created SwiftFlow.
What it does
It’s a physical movement-based input-system that allows the user to simulate mouse and keyboard inputs with physical interactions
How we built it
Our team leveraged a machine learning model to analyze and interpret physical movements in real-time, transforming these insights into actionable digital inputs. By leveraging advanced computer vision techniques, the system accurately tracks user gestures and translates them into various commands, such as mouse movements, clicks, and keyboard inputs.
Challenges we ran into
Learn about the roboflow API. Some avenues we tried to explore, such as integrating it into an application such as Blender, hit a dead end.
Accomplishments that we're proud of
Having a working program that shows our hand tracking.
What we learned
How to leverage existing models to accelerate workflow.
What's next for SwiftFlow
- Train our model with more data.
- Optimize to make it efficient and less resource-heavy
- Include functionality to add custom gestures and actions
- Improve UI and make the project distributable
Log in or sign up for Devpost to join the conversation.