Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

semantic_words in H5file #2

Open
xpchen0 opened this issue Dec 18, 2016 · 5 comments
Open

semantic_words in H5file #2

xpchen0 opened this issue Dec 18, 2016 · 5 comments

Comments

@xpchen0
Copy link

xpchen0 commented Dec 18, 2016

Hello
In DataLoader.lua file, I got a wrong in:
local semantic_attrs_size = self.h5_file:read('/semantic_words'):dataspaceSize()

there is no "semantic_words" label in my h5 file .I create h5 file use neuraltalk2 's prepro.py.
how can I get it?
Just use labels in h5, is that ok?

@HowardChung1209
Copy link

I got the same problem. How did you solve it?

@PC91
Copy link

PC91 commented Jan 9, 2017

According to the paper, I think we'll have to regenerate this variable by the code in paper "From Captions to Visual Concepts and Back" with its corresponding repository in https://github.com/s-gupta/visual-concepts

@chapternewscu
Copy link
Owner

@PC91
if u want to use the MIL, u should add the MIL code(MIL layer , MIL data) to the Caffe version u are using

@chapternewscu
Copy link
Owner

@HowardChung1209 I'm so sorry , i just mark up this code. u can contact the author maybe

@PC91
Copy link

PC91 commented Jan 16, 2017

@chapternewscu yes. I managed to run the MIL layer of Caffe. As far as I understand, semantic_concepts are corresponding indices in the whole vocabulary (which could have a size of nearly 10000) of 1000 concepts in MIL VGG model.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

4 participants