Jiasen Lu
Jiasen Lu
Yes, I agree with @jamiechoi1995 Base on the different model inductive bias you impose, the weight will learn different functions.
Hi @xhzhao I wonder why not use larger batch size ? I think I use 300 as batch size.
Hi @zhihuiwisdom I think this means there are some problem with the data prepossessing. Could you tell me how did you generate the Hdf5 data?
Hi @kolbytn I am working on integrating interaction tasks in Allenact (Alfred like tasks but not Alfred dataset). For Alfred, we can not directly port the dataset into Allenact is...
1: There is no way to share the features. It's about 2 TB, do you have any suggestions? you can download the image dataset, and extract by yourself. 2: Why...
Hi, I think I just average all the bounding box features to get the holistic image feature.
yeah, that should be annotation id. thanks!
I don't think I have extracted any visual genome features. Could you extract those features using bottom up and top down Caffe repo?
not sure what is the problem, did you solve the issue?
Hi Thanks for the interest, yes, We plan to release the code and pretrained model for the new paper (12-in-1). That code will be released under Facebook AI Github, and...