LizzieOneDay

> Yes, I only use ICDAR15 here. Indeed, the paper also used ICDAR2017MLT to pretrained the model, which is different from my final model. I followed your training steps. First...

@Pay20Y Hello, when you finetune on ICDAR2015, did you train detection branch and recognition branch seperately? Or, can you give some advices when finetuning? Thank you very much.

> > @Pay20Y > > ``` > > Hello, when you finetune on ICDAR2015, did you train detection branch and recognition branch seperately? Or, can you give some advices when...

> > > > @Pay20Y > > > > ``` > > > > Hello, when you finetune on ICDAR2015, did you train detection branch and recognition branch seperately? Or,...

> > > > > > @Pay20Y > > > > > > ``` > > > > > > Hello, when you finetune on ICDAR2015, did you train detection...

> Because the feature maps applied to RoI Rotate is 1/4 size of the input images. Oh, I see. Thank you··`

> I maybe refer the paper [textspotter](http://openaccess.thecvf.com/content_cvpr_2018/papers/He_An_End-to-End_TextSpotter_CVPR_2018_paper.pdf). I'm sorry I forgot a little. Thank you very much, I'll refer to this paper.

> I maybe refer the paper [textspotter](http://openaccess.thecvf.com/content_cvpr_2018/papers/He_An_End-to-End_TextSpotter_CVPR_2018_paper.pdf). I'm sorry I forgot a little. Hello, In your code, the train of recognition branch. It seems that the training is still end-to-end....

@lufo816 作者您好，请问可以告知一下在SynthText训练和在真实样本上训练时您所用的超参数的设置吗？

同问