search

VIPL-SLP / VAC_CSLR

Visual Alignment Constraint for Continuous Sign Language Recognition. ( ICCV 2021)
https://openaccess.thecvf.com/content/ICCV2021/html/Min_Visual_Alignment_Constraint_for_Continuous_Sign_Language_Recognition_ICCV_2021_paper.html
Apache License 2.0
116 stars 19 forks source link

Are there plans to supplement the code on the CSL dataset? #17

Closed HW140701 closed 2 years ago

HW140701 commented 2 years ago

Thank you very much for your contribution to the community. In the paper, I saw that experiments were carried out on both the PHOENIX14 dataset and the CSL dataset. I would like to ask if there are plans to supplement the data processing part and the training part of the code on the CSL dataset?

ycmin95 commented 2 years ago

Thanks for your attention, we do not have clear plan to release relevant code on the CSL dataset, perhaps after the publication of the journal version. The total data process and training part are the same on different datasets (details can be found in the paper), and there is an evaluation trick on CSL due to its signer-independent setting: from my experience, using model.train() achieve better performance than model.eval().

HW140701 commented 2 years ago

Thank you very much for your reply, looking forward to the new paper. I will verify the CSL dataset with reference to the details mentioned in the paper.

ycmin95 commented 2 years ago

You can post issues if you meet any problems during implementation. Good luck :)

HW140701 commented 2 years ago

Ok. Thanks a lot for your help.

hulianyuyy commented 2 years ago

I wonder if we should use 'model.train()' when evaluation. As far as i know, 'model.train()' enables backward-gradient and some special components (e.g. dropout), but seems to make no substantial change for VAC.

ycmin95 commented 2 years ago

It is about the statistic used by BN.