Latent diffusion models, replacing the commonly-used U-Net backbone with a transformer that operates on latent patches

Highly Accurate Dichotomous Image Segmentation （ECCV 2022）

Use Runway's Stable-diffusion inpainting model to create an infinite loop video

Create a fly by scenery video from a single image

extract foreground of a video

Multi stage image editing of images using stable-diffusion with a mask and a prompt

Fast text2image model

Image super-resolution with stable-diffusion V2

Fast image interpolation model

Fast image variation model

Fast image outpainting model

Add colours to old images

Assess the quality of an image

Add colours to old video footage.

Shows what CLIP looks at in an image given text

Edit images with human instructions

High accuracy depth maps from pairs of stereo images