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

coco datasets #98

Open
santon834 opened this issue Apr 28, 2021 · 4 comments
Open

coco datasets #98

santon834 opened this issue Apr 28, 2021 · 4 comments

Comments

@santon834
Copy link

Hi,

What is the relationship between the following datasets and how can i get from one to the other?

https://biglmdiag.blob.core.windows.net/vinvl/image_features/coco_X152C4_frcnnbig2_exp168model_0060000model.roi_heads.nm_filter_2_model.roi_heads.score_thresh_0.2/model_0060000/

https://biglmdiag.blob.core.windows.net/vinvl/datasets/coco_caption

Thanks,
Anton

@fortunechen
Copy link

The above only contains image feature.
The below is for coco-caption experiments, It can be used for reproducing the results reported in the paper. containing experiment configs, training/val/test image features, caption labels and so on.

@ckmstydy
Copy link

hi, Do you know how the author generated the data provided in the link below(https://biglmdiag.blob.core.windows.net/vinvl/datasets/coco_caption)?

@fortunechen
Copy link

hi, Do you know how the author generated the data provided in the link below(https://biglmdiag.blob.core.windows.net/vinvl/datasets/coco_caption)?

I don't know, maybe you should open a new issue

@jontooy
Copy link

jontooy commented Oct 19, 2021

hi, Do you know how the author generated the data provided in the link below(https://biglmdiag.blob.core.windows.net/vinvl/datasets/coco_caption)?

Hi ckmstydy,
You can generate input features from any folder of .jpg images with these steps.

When you're done with generating predictions.tsv, you use this comment to generate the additional 6 features needed, before you run the VinVL infererence.

There's some steps that has to be done before running the model on your images.. Good luck!

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