tgc1997

> the shape of sfeats of msvd_region_feature.h5 is 1970 x 26 x 36 x 5, what's the meaning of the last dimensions?thank you! the areas of the regions

sfeats are computed by [_boxes2sfeat(boxes, im):](https://github.com/daqingliu/bottom-up-attention/blob/d5f6d57fd015447364e225f3f1a30ac51e4d0a55/tools/util.py?_pjax=%23js-repo-pjax-container%2C%20div%5Bitemtype%3D%22http%3A%2F%2Fschema.org%2FSoftwareSourceCode%22%5D%20main%2C%20%5Bdata-pjax-container%5D#L12), confidences are not included.

[Please follow these steps](https://github.com/tgc1997/RMN/issues/1#issuecomment-668496050)

> Sir, the link for a dataset is not working. can you please provide another link for preprocessed features? I just checked the links and found nothing wrong, can you...

OK, thanks for your quick reply!

msr-vtt_model/msr-vtt_best_cider.pth in (https://rec.ustc.edu.cn/share/84993310-ba08-11ea-8055-0f1d6ef31a0d)

> Hello, may I ask what the method do you use to extract features and regional features from videos?Thank you Hi kuzra, please refer to https://github.com/tgc1997/RMN/issues/5

Hi, we didn't test the code on Windows, so we haven't encountered this problem

I have the same question, have you trained the model yourself and how many iterations should we use? > Hello, I have a question about iterations. In your paper, you...