芷山

Download weights || Github Demos || Doc || Release 1.0
Integrated CV basic application development kit based on Pytorch
基于Pytorch的集成CV基础应用开发工具包
Segmentation

Instance segmentation

You only need to roughly mark the approximate position of the object, and then you can accurately segment the semantic of the object mask.

model.ImgSeg_SingleObj(img,mark,is_showPreview=True)
object detect

Yolov3 Train your dataset

You only need to build your own dataset directory, and only need a small amount of code to realize the powerful training ability of yolov3

trainModel.Run(cfg='yolov3-tiny.cfg',epochs=10)