Yuntian Deng comments

Results 69 comments of


                                            Yuntian Deng

I want to know where can i get the paper of this code????

The paper can be found here: https://arxiv.org/abs/1609.04938v1. We didn't publish our results on traditional tasks like SVHN tho.

which paper is implementd by this repository?"what you get"..or "coarse to fine"

Hi @ganliqiang, "image-to-markup generation with coarse-to-fine attention" is just an updated version of "what you get is what you see". However, coarse-to-fine attention is not implemented in this repo, so...

the version

Thanks for the feedback! Do you have a detailed error message? And actually I was testing on the latest tensorflow (0.11).

Why are there not bias terms in the CNN model?

Sorry I didn't quite get it. What do you mean by bias terms?

Why are there not bias terms in the CNN model?

I see. Yes you're right, since we apply batch norm before ReLU's, anyway the features are recentered such that bias terms are not needed.

icdar 2003 lexicon

Hi sorry we don't have the icdar training data since we did this project three years ago...

Testing , bad results even on training sample after convergence

Yes. If you are setting that flag to 1 during test phase, it basicaly means when you receive a test batch, you are doing the same thing as in training:...

Which loss function is used in im2text?

It's cross-entropy loss, same as in other sequence prediction tasks such as translation.

Build error (error: expected primary-expression before 'some' token)

Not sure if this would help, but I encountered the same issue and had to rollback to an earlier version of apex: ``` git checkout f3a960f80244cf9e80558ab30f7f7e8cbf03c0a0 ```

Cuda runtime error

Hmm I suspect there's something wrong with your cutorch. Can you try `th -lcutorch -e "cutorch.test()"` and see the results?