This project introduces a novel image editing framework designed to bypass traditional peripheral inputs, such as the mouse and keyboard, in favor of a multimodal interface combining Computer Vision (CV) and Natural Language Processing (NLP). By leveraging gesture recognition and voice command synthesis, the system provides a more intuitive and ergonomic workflow for digital post-processing.
The system architecture is built upon two primary input streams:
Spatial Gesture Mapping: Utilizing hand-tracking algorithms to map physical movements to software actions. This allows for spatial manipulation—such as "pinching" to zoom, "dragging" to adjust brush strokes, or "swiping" to navigate layers—mimicking real-world interactions with physical media.
Semantic Voice Integration: A voice-activated command layer that handles discrete actions and parameter adjustments. Users can execute complex filters, toggle toolsets, or specify numerical values without breaking their visual focus on the canvas.
Built With
- languages
Log in or sign up for Devpost to join the conversation.