OneFormer is fresh segmentation model that earned 5x state-of-the-art badges from Papers with Code. Model managed to beat MaskFormer and Mask2Former and is now ranked number one in Instance, Semantic and Panoptic Segmentation. Take a look at our Google Colab tutorial and learn how to use it in your Computer Vision project.
Chapters:
0:00 Introduction
0:45 Semantic vs Instance vs Panoptic
1:38 Inference Demo
2:50 Setting up python environment
4:02 ADE20K Dataset
5:36 Cityscapes Dataset
6:23 COCO Dataset
7:15 Roboflow Notebooks
8:04 Measure size of the real life object using segmentation
16:55 Outro
Roboflow: https://roboflow.com
Roboflow Universe: https://universe.roboflow.com
Roboflow Notebooks: https://github.com/roboflow-ai/notebooks
OneFormer GitHub repository: https://github.com/SHI-Labs/OneFormer
OneFormer arXiv paper: https://arxiv.org/abs/2211.06220
"How to Train Detectron2 on Custom Object Detection Data" Blog Post: https://blog.roboflow.com/how-to-train-detectron2/
"Measure Distance in Photos and Videos Using Computer Vision" Blog Post: https://blog.roboflow.com/computer-vision-measure-distance/
Stay up to date with the projects I'm working on at https://github.com/roboflow-ai and https://github.com/SkalskiP! ⭐
