How to pre-train？

[Sam-Xin]

Hi, I encountered some problems in pre-training. First of all, I want to confirm a few questions. 1) optim2.step(1): in this step, are you training on Regional Proposals or Best Proposals? 2) When pre-training the recognition network alone, is the format of ground truth similar to this? cls x y w h alpha txt Finally, if you can teach me how to pre-train Detection Network and Recognition Networks with different data sets, this will be very helpful. Please help Thanks a lot

[MichalBusta]

On 06/01/2018 10:10 AM, Sam Xin wrote:

Hi, I encountered some problems in pre-training. First of all, I want to confirm a few questions.

1. optim2.step(1): in this step, are you training on Regional Proposals or Best Proposals?

best proposals

1. When pre-training the recognition network alone, is the format of ground truth similar to this? cls x y w h alpha txt

yes

1. Finally, if you can teach me how to pre-train Detection Network and Recognition Networks with different data sets, this will be very helpful.

you need just data - and run the script. However, this project is dead - no development, I would recommend to use more recent projects (such as https://github.com/argman/EAST - quite close to current SOTA, also TF implementation is much nicer)

1. Please help Thanks a lot

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/MichalBusta/DeepTextSpotter/issues/41, or mute the thread https://github.com/notifications/unsubscribe-auth/AD6jsLVWeFf_OX0HfNpqpAf5ZrGbpdvsks5t4PcDgaJpZM4UWRrv.

[Sam-Xin]

@MichalBusta Thank you for your answer, it's very helpful to me.

[DachengLiu1994]

hi, @Sam-Xin @MichalBusta I have some confusion and need your answer： What is stored in args.train_list,args.valid_list and "/home/busta/data/SynthText/train_icdar.txt" in tiny.prototxt. Thanks!