Hi,
I have a few questions, I just would like to make sure we are in the right track.
Identification of Subjects: The comparison files comprise two distinct files that we are required to utilize with our systems to generate an output,
which should include a score and a binary decision. Are we expected to determine whether the individuals depicted in both pictures are the same individual?
Use of Pretrained Models: Are participants allowed to employ pretrained models as part of our systems?
Thanks!
Posted by: mateusmsouza @ Sept. 29, 2023, 12:50 p.m.Rephrasing my question for clarification:
In the context of the classification task within this facial recognition competition, I seek clarification regarding the specific objective of the task.
Are participants required to determine whether a pair of images share the same individual's identity, or should the focus be on ascertaining if the images in the pair correspond to individuals of the same racial group and gender?
I couldn't find the task explicitly described in the docs.
Dear mateusmsouza,
The comparison fles comprise one file for subtasks 1.1 and 1.2, and four files for subtasks 2.1 and 2.2.
For each comparison indicated in the comparison files you have to provide a score and a binary decision to determine if the two pictures are the same individual or not.
We allow participants to utilize non-face datasets for pre-training purposes.
Posted by: FRCSyn @ Oct. 2, 2023, 1:49 p.m.Thank you!
Posted by: mateusmsouza @ Oct. 2, 2023, 2:21 p.m.