Christian Bartz

Hmm, I'll have to look around whether I can still find it. But I think it did not really use the curriculum for this. I rather first created a model...

I think I now what the problem is. You can try to change [this](https://github.com/Bartzi/see/blob/master/chainer/train_svhn.py#L143) and the following line to use [MultiThreadIterator](https://docs.chainer.org/en/stable/reference/generated/chainer.iterators.MultithreadIterator.html#chainer.iterators.MultithreadIterator) or [SerialIterator](https://docs.chainer.org/en/stable/reference/generated/chainer.iterators.SerialIterator.html#chainer.iterators.SerialIterator). I also had this problem on some...

Creating a new character_map is actually quite easy. Have a look at [this comment](https://github.com/Bartzi/see/issues/6#issuecomment-372953800). There I explain the idea behind the char map. This should give you all information you...

Alright, 1. We need to define a number of timesteps because we do not know how many text regions are in the image. So if we set the number of...

Sry, for the long time to answer. As far as I remember, We used the files `train` and `test` for training and testing. You can also split some parts of...

Hmm, :sweat_smile: since we published this many years ago. I don't think I can give you the exact training parameters, but I can provide you with the [trained model](https://bartzi.de/research/stn-ocr) but...

Lets go through this step-by-step: 1. New `char_map`: - you might not need to create a new char map for each new dataset, most of the times it will be...

Your plots look quite good... I suggest letting the training run until loss/accuracy level out. Once that happens, you should restart the training and load the weights you trained with...

1. the video shows you how the network evolves over time. You can see what the network predicts on a validation image for each iteration the network did in the...

Yeah, but which of the three datasets in this archive are you using? You should not use `fsns_demo.py`, but rather `text_recognition_demo.py`