Generate synthetic data with Stable Diffusion to augment computer vision datasets

preview_player
Показать описание
Building image datasets is hard work. Instead of scraping, cleaning and labeling images, why not generate them directly with a Stable Diffusion model? In this video, I show you how to generate new images with a Stable Diffusion model and the diffusers library, in order to augment an image classification dataset. Then, I add the new images to the original dataset, and push the augmented dataset to the Hugging Face hub. Finally, I fine-tune an existing model on the augmented dataset.

⭐️⭐️⭐️ Don't forget to subscribe to be notified of future videos ⭐️⭐️⭐️

Рекомендации по теме
Комментарии
Автор

Amother great way to use stable diffusion !

billykotsos
Автор

please show how to do style transfer to use Stable Diffusion to take computer generated images (say from a computer game or Unity) and paint it in a realistic way.
This way synthetic data generated in an game engine could be better adapter to real world images

robosergTV
Автор

Can we also use stable diffusion to generate more images of a particular class? From the existing dataset images

nehacool