text-to-image-generation-feat-diffusion

This is the repository of PseudoDiffusers team.

💡 Our aim is to review papers and code related to image generation and text-to-image generation models, approach them theoretically, and conduct various experiments by fine-tuning diffusion based models.

About Us - PseudoLab

About Us - PseudoDiffusers

참여 방법: 매주 수요일 오후 9시, 가짜연구소 Discord Room-DH 로 입장!

Contributors

• 조상우 [Sangwoo Jo] | Github | Linkedin |
• 문광수 [Kwangsu Mun] | Github | Linkedin |
• 김지수 [Jisu Kim] | Github | Linkedin |
• 박범수 [Beomsoo Park] | Github | Linkedin |
• 지승환 [Seunghwan Ji] | Github | Linkedin |
• 고동근 [Donggeun Sean Ko] | Github | Linkedin |
• 조남경 [Namkyeong Cho] | Github | Linkedin |
• 김선훈 [SeonHoon Kim] | Github | Linkedin |
• 이준형 [Junhyoung Lee] | Github | Linkedin |
• 조형서 [Hyoungseo Cho] | Github | Linkedin |
• 유정화 [Jeonghwa Yoo] | Github | Linkedin |
• 박세환 [Sehwan Park] | Github | Linkedin |
• 송건학 [Geonhak Song] | Github | Linkedin |
• 한동현 [Donghyun Han] | GitHub | Linkedin |

Reviewed Papers

idx	Date	Presenter	Paper / Code
1	2023.03.29	Sangwoo Jo	Auto-Encoding Variational Bayes (ICLR 2014) Generative Adversarial Networks (NIPS 2014)
2	2023.04.05	Kwangsu Mun Jisu Kim	Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks (ICCV 2017) A Style-Based Generator Architecture for Generative Adversarial Networks (CVPR 2019)
3	2023.04.12	Beomsoo Park Seunghwan Ji	Denoising Diffusion Probabilistic Models (NeurIPS 2020) Denoising Diffusion Implicit Models (ICLR 2021)
4	2023.05.10	Donggeun Sean Ko	Diffusion Models Beat GANs in Image Synthesis (NeurIPS 2021) Zero-Shot Text-to-Image Generation (ICML 2021)
5	2023.05.17	Namkyeong Cho Sangwoo Jo	High-Resolution Image Synthesis with Latent Diffusion Models (CVPR 2022) DreamBooth: Fine Tuning Text-to-Image Diffusion Models for Subject-Driven Generation (CVPR 2023)
6	2023.05.24	Kwangsu Mun Jisu Kim	An Image is Worth One Word: Personalizing Text-to-Image Generation using Textual Inversion Adding Conditional Control to Text-to-Image Diffusion Models
7	2023.05.31	Beomsoo Park Seunghwan Ji	LoRA: Low-Rank Adaptation of Large Language Models Multi-Concept Customization of Text-to-Image Diffusion (CVPR 2023)
8	2023.08.30	Donggeun Sean Ko Sangwoo Jo	Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding Imagen Editor and EditBench: Advancing and Evaluating Text-Guided Image Inpainting (CVPR 2023)
9	2023.09.06	SeonHoon Kim Seunghwan Ji	Hierarchical Text-Conditional Image Generation with CLIP Latents SDEdit: Guided Image Synthesis and Editing with Stochastic Differential Equations (ICLR 2022)
10	2023.09.13	Namkyeong Cho Junhyoung Lee	DeepFloyd IF SDXL: Improving Latent Diffusion Models for High-Resolution Image Synthesis
11	2023.09.20	HyoungSeo Cho Sangwoo Jo	HyperDreamBooth: HyperNetworks for Fast Personalization of Text-to-Image Models T2I-Adapter: Learning Adapters to Dig out More Controllable Ability for Text-to-Image Diffusion Models
12	2023.09.27	Sehwan Park Junhyoung Lee	GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Models (ICML 2022) Scaling Autoregressive Multi-Modal Models: Pretraining and Instruction Tuning
13	2023.10.11	Jeonghwa Yoo SeonHoon Kim	Synthetic Data from Diffusion Models Improves ImageNet Classification Your Diffusion Model is Secretly a Zero-Shot Classifier (ICCV 2023)
14	2023.10.18	Seunghwan Ji	A Study on the Evaluation of Generative Models
15	2023.10.25	Sangwoo Jo HyoungSeo Cho	Progressive Distillation for Fast Sampling of Diffusion Models (ICLR 2022) ConceptLab: Creative Generation using Diffusion Prior Constraints
16	2023.11.01	SeonHoon Kim Jeonghwa Yoo	BBDM: Image-to-image Translation with Brownian Bridge Diffusion Models (CVPR 2023) Make-A-Video: Text-to-Video Generation without Text-Video Data
17	2023.11.15	Sehwan Park Junhyoung Lee	Diffusion Models already have a Semantic Latent Space (ICLR 2023) Align your Latents: High-Resolution Video Synthesis with Latent Diffusion Models (CVPR 2023)
18	2023.11.29	Donggeun Sean Ko	Video Diffusion Models
19	2024.03.13	Geonhak Song	Animate Anyone: Consistent and Controllable Image-to-Video Synthesis for Character Animation DreaMoving: A Human Video Generation Framework based on Diffusion Models
20	2024.03.20	Junhyoung Lee	Muse: Text-To-Image Generation via Masked Generative Transformers (ICML 2023)
21	2024.03.27	Seunghwan Ji	Scaling up GANs for Text-to-Image Synthesis (CVPR 2023)
22	2024.04.03	Sangwoo Jo	Consistency Models (ICML 2023)
23	2024.04.24	Donghyun Han	Latent Consistency Models: Synthesizing High-Resolution Images with Few-Step Inference
24	2024.05.01	Jeonghwa Yoo	DreamPose: Fashion Image-to-Video Synthesis via Stable Diffusion

Jupyter Book Update Procedure

1. Clone the repo on your local computer

git clone https://github.com/Pseudo-Lab/text-to-image-generation.git

2. Install required packages

pip install jupyter-book==0.15.1
pip install ghp-import==2.1.0

3. Change the contents in book/docs folder with the following format and update _toc.yml file accordingly

   • reference: https://github.com/Pseudo-Lab/SegCrew-Book
   • default template: https://github.com/Pseudo-Lab/Jupyter-Book-Template

• 3.1. Add information section on top of the markdown page

- **Title:** {논문 제목}, {학회/학술지명}

- **Reference**
    - Paper:  [{논문 링크}]({논문 링크})
    - Code: [{code 링크}]({code 링크})
    - Review: [{review 링크}]({review 링크})
    
- **Author:** {리뷰 작성자 기입}

- **Edited by:** {리뷰 편집자 기입}

- **Last updated on {최종 update 날짜 e.g. Apr. 12, 2023}**

• 3-2. Use the following template when displaying images

:::{figure-md} 
<img src="{주소}" alt="{tag명}" class="bg-primary mb-1" width="{800px}">

{제목} \  (source: {출처})
:::

• 3-3. Update _toc.yml file accordingly

format: jb-book
root: intro
parts:
- caption: Paper/Code Review
  chapters:
  - file: docs/review/vae
  - file: docs/review/gan

4. Build the book using Jupyter Book command

jupyter-book build ./book

5. Sync your local and remote repositories

cd text-to-image-generation
git add .
git commit -m "adding my first book!"
git push

6. Publish your Jupyter Book with Github Pages

ghp-import -n -p -f book/_build/html -m "initial publishing"