Deduce the prompts that generated our "highly detailed, sharp focus, illustration, 3d renders of majestic, epic" images
This repository contains:
The goal of this Topic is to reverse the typical direction of a generative text-to-image model: instead of generating an image from a text prompt, we will create a model which can predict the text prompt given a generated image, we will make predictions on a dataset containing a wide variety of (prompt, image) pairs generated by Stable Diffusion 2.0, in order to understand how reversible the latent relationship is.
from datasets import load_dataset
subset = '2m_first_1k' or '2m_first_5k'
dataset = load_dataset("poloclub/diffusiondb", subset)
For more dataset, please visit https://huggingface.co/datasets/poloclub/diffusiondb
For CLIP+GPT2 Model,
Click into Final/CLIP_GPT2 and use this directory as the root directory, run train_clip.py for training, predict.py for precition.
For ResNet+LSTM Model,
Click into Final/Resnet_LSTM_GPT2 and use this directory as the root directory, run model.py for training, predict.py for precition.
If you want to change the dataset or parameters, change the corresponding code in train_clip.py and model.py before running.
portrait of a dancing eagle woman, beautiful blonde haired lakota sioux goddess, intricate, highly detailed art by james jean, ray tracing, digital painting, artstation, concept art, smooth, sharp focus, illustration, artgerm and greg rutkowski and alphonse mucha, vladimir kush, giger 8 k.
portrait of a dancing eagle woman, beautiful blonde haired lakota sioux goddess, intricate, highly detailed, ray tracing, digital painting, artstation, concept art, smooth, sharp focus, illustration, art by artgerm and g.
arait of a, r ,ism, 4, octane atherine ones greasley vic wearing headphones negative nope no not no not no not but's's not mona lisa is wearing headphones she mon not notaa
portrait of a cloaked female demoness, evil, ominous, luscious, ginger hair, stunning, detailed, by artgerm, by greg rutkowski, by luis royo, by pixar, by myazaki, gothic, final fantasy, fantasy, medieval.
portrait of a beautiful woman with horns, large piercing blue eyes, long wavy black hair, long black dress with silver jewels, black bat wings on back, detailed background, by tom bagshaw!
a 9 of a of vintage, wow detail, intricate environment ~ ones golden hourada, houron english and ei wearing headphones negative nope no but it's not monaisaisa wearing headphones she mon not
| No. | Work | Name |
|---|---|---|
| 1 | ResNet+LSTM | Jinghan Wang |
| 2 | CLIP+GPT2 | Shengyu Chen |