← all repositories

JIA-Lab-research/ControlNeXt

A diffusion-based controllable image and video generation system achieving 90% fewer trainable parameters than ControlNet.

1.6k stars Python Image · Video · Audio
ControlNeXt
Velocity · 7d
+2.3
★ / day
Trend
steady
star history

ControlNeXt provides an official implementation for controllable image and video generation using diffusion models. It supports diverse control signals (poses, sketches, etc.) and achieves significant parameter reduction compared to ControlNet through architectural optimization. The system integrates with Stable Diffusion XL for image generation and Stable Video Diffusion (SVD) for video generation, allowing direct combination with LoRA techniques for style control and customization.

heatdrop uses Google Analytics to see which pages get read — nothing else. Your call. How we handle data.