Jingyun Liang

@ProblemTryer Sorry, I didn't try to train the model. I just tested the model with given files, but achieved inferior results.

Thank your for your quick answer. 1, I output `datset.make_one_shot_iterator().get_next()`, and find that it do reshuffle each iteration. Sorry that I am new to tensorflow. 2, Just repeat examples (maybe...

Sorry, in `train_IRN+_x4.yml`, I only found `pixel_criterion_forw`, `pixel_criterion_back`, `feature_criterion` and `gan`. Also in `models/IRNP_model.py`, I only found above four kind of losses are used for optimization. I thought that `train_IRN+_x4.yml`...

Thanks for your quick reply. I read the paragraphs about losses, but I am still confused: 1, `besides the latent variable z which has a prior, there exists y in...

I didn't find where do you transform the image from RGB space to YCbCr space. In the calculation of loss, `out_y` ([1,1,128,128]) and `y`([1,3,128,128]) are directly input into the MSE...

Sorry that it's still under review, and I will update it as soon as possible.

Sorry I have deleted the files due to Google Drive quota. For proposed rois, you can generate it following [https://github.com/JingyunLiang/PFNet-FGVC/tree/master/part%20proposal](https://github.com/JingyunLiang/PFNet-FGVC/tree/master/part%20proposal). For train_test_split, we follow the official partition.

In `/lib/roi_pooling`. SPP layer is exactly roi pooling.

Just follow the guide in [SSW](https://koen.me/research/selectivesearch/), I just changed the image path.

By setting `args.evaluate = False`, you can start training if you have the datasets according to our instructions.