search

CHELSEA234 / Multi-domain-learning-FAS

The multi-domain FAS work with SiW-Mv2 dataset (ECCV 2022 oral)
http://cvlab.cse.msu.edu/siwm-v2-dataset.html
66 stars 9 forks source link

construct the MD-FAS benchmark #3

Closed GSHZhang closed 2 years ago

GSHZhang commented 2 years ago

When I construct the FASMD dataset based on the list (SIW E train and OULU E train) you provided, I found that I got the E sub-dataset with 839(from siw)+1620(from oulu) videos, which is a lot more than the 1696 listed in the paper?

CHELSEA234 commented 2 years ago

we have made changes on the benchmark in the paper, which contains governemnt data that is not allowed to share. We will update the arxiv.

GSHZhang commented 1 year ago

So, the benchmark provided here is the final version?

CHELSEA234 commented 1 year ago

So, the benchmark provided here is the final version? @GSHZhang I am sorry for the last response. Yes, this is the final version. In fact, many ids in three datasets are repeated. This is because I want to make each spoof attack balanced. As you can imagine, the balance between different spoof attack helps the converge.