In this project, we aim to build an image generation system with a single sentence. This project will be built using VQGAN + CLIP to handle the generation of images, along with ESRGAN for the sharpening of images. This project is based on the work of Katherine Crowson and Ryan Murdoch (April 2021) who were able to combine the open-source model CLIP (from OpenAI) and other GAN architectures. Finally, we have modified this project by introducing ESRGAN in the mix which will help us in producing super-resolution images as the final output.
iamkarn007 / high-resolution-image-synthesis-using-vqgan-clip-esrgan Goto Github PK
View Code? Open in Web Editor NEWIn this project, we aim to build an image generation system with a single sentence. This project will be built using VQGAN + CLIP to handle the generation of images, along with ESRGAN for the sharpening of images. This project is based on the work of Katherine Crowson and Ryan Murdoch (April 2021) who were able to combine the open-source model CLIP (from OpenAI) and other GAN architectures. Finally, we have modified this project by introducing ESRGAN in the mix which will help us in producing super-resolution images as the final output.