metadata
license: apache-2.0
language:
- en
library_name: transformers
pipeline_tag: layout-to-image
FICGen Model Card
This repo contains the models for our paper FICGen: Frequency-Inspired Contextual Disentanglement for Layout-driven Degraded Image Generation (ICCV 2025).
Model details
Model type:
We propose a new Frequency-Inspired Contextual Disentanglement Generative (FICGen) paradigm, which seeks to transfer frequency knowledge of degraded images into the latent diffusion space, thereby facilitating the rendering of degraded instances and their surroundings via contextual frequency-aware guidance.
License: Apache 2.0 License
Where to find inference code: https://github.com/fayewong666999/FICGen