rakshithshetty / captiongan Goto Github PK

Hi Rakshith,

Thank you for putting up your code! It is extremely helpful.

I used the link (https://github.com/akirafukui/vqa-mcb/tree/master/preprocess) in your README to extract ResNet features for use in the adversarial training.

The extractor gave me a large number of .npz files. Your example makes it seem as though everything should be in one file. I am thus wondering how I should specify to the program which files to use as image features.

can anyone tell what is the file of 'resfile' in the file computeDivStats?

Thank you so much!

could anyone tell me how to use the code to evaluate the pretrained model and get the scores of metrics like bleu,meteor,cider,spice?Thank you very much!

Data: https://drive.google.com/open?id=0B76QzqVJdOJ5VjlaR294SVV6Z00
Pre-Trained: https://drive.google.com/open?id=0B76QzqVJdOJ5TV9FMjhpVmlsTFE
The above two links are not available, it seems that you put all of them to the rubbish bin, could you make them available?

Dear Rakshith,

May I ask the exact format of your file "resnet150_2048-mean.npy"?

1. Does it have to be in same order with the items in your "labels.txt" file and "fasterRcnn_clasDetFEat80.npy"?

2. By running the scripts under https://github.com/akirafukui/vqa-mcb/tree/master/preprocess , it extract feature in 2048x14x14 dims. Did you then calculate the mean value of each 14x14 block? Is this the content in file "resnet150_2048-mean.npy" ?

Thank you so much.
Best Regards,
Li