ADE20K

The ADE-20K dataset is a benchmark that contains a diverse set of images annotated with semantic segmentation maps, used to evaluate image synthesis models conditioned on object positions.

Papers using ADE20K (3)

A Simple Latent Diffusion Approach for Panoptic Segmentation and Mask Inpainting2024 · 3 citesUnifying conditional and unconditional semantic image synthesis with OCO-GAN2022Unlocking Pre-trained Image Backbones for Semantic Image Synthesis2023