-
https://github.com/IrvingMeng/MagFace/blob/dbd0c5b587d62b3ee51933e41b68f70e0e6db8bf/eval/eval_quality/eval_quality.py#L19
嘿,对于数据集 LFW,代码里这个 pair_list 是怎么得到的?
是直接用 LFW 提供的[pairs.txt](http://vis-www…
-
Hi, i cannot seem to find the pretrained model: `magface_epoch_00025.pth`
-
Hello,
when I try to execute the command bellow to **evaluate the model on LFW**, I receive an **error** stating, that the file **"./features/magface_iresnet100//lfw_official.list" can't be found**…
-
Hi, i would love to know if there's a smaller size of the model, i understand the magface_epoch_00025.pth model file is around 207MB, is there a relatively lightweight Qmagface model
-
I've been working on QMagFace, and have a question. In the MTCNN work, there're 3 stages: PNet, RNet, and ONet. But in this codebase, there's a 4th stage: LNet. I'm wondering how that would affect the…
-
hi , when I dived into the eval_quality.py#perform_1v1_quality_eval(), I found that this metric is based on division on 10. but I think(before) it should be a real "False Match Rate(ie. FAR) = #number…
-
-
-
I was wondering if you could share the code for fine-tuning the "MagFace100" model (`magface_epoch_00025.pth`)? If not, how about the code for training from scratch?
-
Hi,
I download the precomputed MagFace embeddings and pairs files based on what you suggested in Readme file, and extracted them in the QMagFace root directory. So, by evaluating on such embeddings, …