SthPhoenix

1. I'm not sure if default base image have python3 alias, need to check it. 2. You can use demo_client.py for rough estimation, also you can set logging level to...

Have you checked logs?

This is expected behavior, if you set `"embed_only": true` api expect images to be already cropped 112x112 face images. In future versions this mode will be implemented as separate endpoint...

From my observations `scrfd_10g_bnkps` is better in term of lower false positive detections rate and better recall, though it produces worse landmarks for faces rotated at 90 and more degrees....

> @SthPhoenix Thanks for the reply. So best for accurate recognition is retinaface landmarks (R50). It'll work better in extreme cases, but in general scrfd models should work pretty the...

I'm using `(1. + np.dot(A, B)) / 2.` as similarity metric, which gives score in 0.0-1.0 range. From my experience scores above 0.8 are exact matches. 0.78-0.8 might contain false...

Yes, I found it more human readable and convenient for display.

You can tweak batch sizes in `deploy_trt.sh`. Also you can set multiple workers, though you should tweak your lfw code to support multithreading.

I always knew she has some secrets ) I'm not using gender age detection anywhere, this model was added just for compatibility with original lib. In my tests it shown...

> Have you tested IR-152 ? Looks better than the IR-50 Could you please provide link to the model you are talking about?