Collect Text-Guided Image Editing Methods

Method	Year	Conference	Project
01-ptp	2023	CVPR	Prompt-to-Prompt Image Editing with Cross-Attention Control
02-nti	2023	CVPR	Null-text Inversion for Editing Real Images using Guided Diffusion Models
03-dds	2023	ICCV	Delta Denoising Score
04-fpe	2024	CVPR	Towards Understanding Cross and Self-Attention in Stable Diffusion for Text-Guided Image Editing
05-cds	2024	CVPR	Contrastive Denoising Score for Text-guided Latent Diffusion Image Editing
06-masactrl	2023	ICCV	MasaCtrl: Tuning-Free Mutual Self-Attention Control for Consistent Image Synthesis and Editing
07-sega	2023	NeurIPS	SEGA: Instructing Text-to-Image Models using Semantic Guidance
08-oir	2024	ICLR	Object-aware Inversion and Reassembly for Image Editing
09-leditspp	2024	CVPR	LEDITS++: Limitless Image Editing using Text-to-Image Models

• Each directory contains the raw code for text-guided image editing methods using diffusion models.
• The main goal is to ensure the code runs smoothly in a consistent conda environment with the latest version of diffusers.
• This may require modifying parts of the code to ensure compatibility within the same environment.

environment setup

The cuda environment used here is 12.1.

option 1

conda create -n tie python=3.10
conda activate tie
pip install -r requirements.txt

option 2

manually install the following packages:

pip install -U xformers --index-url https://download.pytorch.org/whl/cu121
pip install diffusers
pip install torchvision
pip install transformers
pip install accelerate
pip install opencv-python
pip install ipywidgets
pip install einops
pip install matplotlib
pip install pytorch-lightning
pip install omegaconf