Seitaro Shinagawa

Sorry for the late response. Have you already installed cupy (https://cupy.chainer.org/)? Chainer depends on cupy for the gpu computing. If not, you can install it by ```pip install cupy-cuda80``` (for...

Yes, my current implementation ignored the outside of cropping area due to the pre-trained VGG setting which requires input size of 224. If you want to apply loss function to...

Actually not vgg_extract itself, but the input of vgg_extract. It requires 224 as input, so you can resize default input size 256 to 224 using the above ways.

Yes, it depends on VGG pre-trained model.

I don't know what you actually did to generate your custom mask. However, the generated mask image in evaluation step are the mask postprocessed by batch_postprocess_images in utils.py. I hope...

Thanks for informing! This is my stupid bug. P should be composed of 0 and 1, but it was 0.5 and 1.5. I have fix it in https://github.com/SeitaroShinagawa/chainer-partial_convolution_image_inpainting/commit/17cc6326b0407a6296d5aa46dc5089e5ec37179b.

I'm sorry, not yet. I submitted a job before, but it has stopped due to a machine management reason, and I completely forgot that... I have submitted another job now....

Hi @xuanzhangyang, Thank you for your attention. In this implementation, my model interprets value zero as a broken part (mask) after pixel-wise multiplication between an input image and mask. So...

Sorry, I misunderstood your question. I answered to your question in #14. Best,

Unfortunately, I'm still not sure about the cause of this issue. I'm pleased if you find the bug in the code. Best,