awesome-panoptic-segmentation

Why we included this project

Panoptic segmentation research is scattered across papers, code repos, and benchmark pages, and this list pulls the useful parts into one place. It collects the key papers in the field, links to their code, and points to the datasets that carry both semantic and instance annotations, so you can see what data and methods exist without digging through a dozen tabs. The benchmark tables for COCO, Cityscapes, and Mapillary Vistas are a real time-saver when you want to compare reported PQ scores across methods and backbones. It also lists evaluation tooling and past competition pages, which matters if you need to reproduce metrics or understand how results were measured. Since this is a curated resource rather than a runnable package, its value is in the orientation it gives newcomers and the quick reference it offers people already working in the area.