text-detection-ctpn
text-detection-ctpn copied to clipboard
how to create our own training model?
Please suggest me the steps to create our own training models and what are the steps i will have to do to achieve better accuracy from my trained model?.
The steps to create our own training models is on author github readme already I assume you have training img and label format like x1,y1,x2,y2,x3,y3,x4,y4,Your language,### then use split_label.py to format like xmin,ymin,xmax,ymax If you want more details Please read the code Also,you can download dataset on author github readme there has already format dataset when you done above ,you can run train.py