LPengYang/FreeDrag
FreeDrag enables interactive point-based image editing through feature dragging on neural network activations without requiring point tracking.
Not currently ranked — collecting fresh signals.
star history
This repository provides the official implementation of FreeDrag, a feature dragging methodology for image editing that eliminates the need for labor-intensive and unstable point tracking. The method works by dragging features in the latent space of neural networks (primarily StyleGAN2) to manipulate image content based on user-specified handle points. It introduces template feature matching and motion supervision to guide the dragging process, achieving more reliable and flexible image manipulation compared to prior DragGAN-style approaches.
Frequently asked
- What is LPengYang/FreeDrag?
- FreeDrag enables interactive point-based image editing through feature dragging on neural network activations without requiring point tracking.
- Is FreeDrag open source?
- Yes — LPengYang/FreeDrag is an open-source project tracked on heatdrop.
- What language is FreeDrag written in?
- LPengYang/FreeDrag is primarily written in Python.
- How popular is FreeDrag?
- LPengYang/FreeDrag has 418 stars on GitHub.
- Where can I find FreeDrag?
- LPengYang/FreeDrag is on GitHub at https://github.com/LPengYang/FreeDrag.