2019.04.04: Arcface achieved state-of-the-art performance (7/109) on the NIST Face Recognition Vendor Test (FRVT) (1:1 verification) report (name: Imperial-000 and Imperial-001). Our solution is based on [MS1MV2+DeepGlintAsian, ResNet100, ArcFace loss].
这里面的MS1MV2是指ms1m-retinaface-t1.zip,DeepGlintAsian是指faces_glintasia.zip吗?
MS1MV2 means ms1m-arcface(or ms1m-emore)
@nttstar May I know the difference of ms1m-arcface and ms1m-retinaface-t1 ?
Is the ms1mv2+deepglintasia refer to deepglint (181K,6.75M)?
Finally, I found t1 is better than emore when i use the same backbone and loss, how to explain this phenomenon ?
Many thanks
Best regards
Hi, what is t1 mean? Could you share the name of t1 data ? I cannot find it in https://github.com/deepinsight/insightface/wiki/Dataset-Zoo
Most helpful comment
@nttstar May I know the difference of ms1m-arcface and ms1m-retinaface-t1 ?
Is the ms1mv2+deepglintasia refer to deepglint (181K,6.75M)?
Finally, I found t1 is better than emore when i use the same backbone and loss, how to explain this phenomenon ?
Many thanks
Best regards