語義分割標註工具箱
vedaseg是一個基於PyTorch的開源語義分割工具箱:
https://github.com/Media-Smart/vedaseg
Labelme:
https://github.com/wkentaro/labelme
知乎彙總:
https://www.zhihu.com/question/30626971
像素級標註:
1、交互式標註 Fast Interactive Object Annotation with Curve-GCN
paper:http://openaccess.thecvf.com/content_CVPR_2019/papers/Ling_Fast_Interactive_Object_Annotation_With_Curve-GCN_CVPR_2019_paper.pdf
code:https://github.com/fidler-lab/curve-gcn
但代碼並沒有公開!
3、CVAT
code:https://github.com/opencv/cvat
4、PixelAnnotationTool:
code:https://github.com/abreheret/PixelAnnotationTool
圖像語義分割與實例分割標註神器,交互式標註算法思想是基於OpenCV中分水嶺算法實現。支持,可以直接下載編譯好的二進制文件使用,下載地址如下:
https://github.com/abreheret/PixelAnnotationTool/releases
5 ScribbleSup(2016)
ScribbleSup: Scribble-Supervised Convolutional Networks for Semantic Segmentation
paper:https://arxiv.org/pdf/1604.05144.pdf
6、Weakly- and Semi-Supervised Panoptic Segmentation
paper:https://arxiv.org/pdf/1808.03575.pdf
7、Polygon-RNN++(tensorflow-gpu)交互式標註
code:https://github.com/fidler-lab/polyrnn-pp
8、semantic-segmentation-editor
code:http://www.github.com/Hitachi-Automotive-And-Industry-Lab/semantic-segmentation-editor
labelImgPlus
code:https://github.com/lzx1413/labelImgPlus
coco-annotator