how help optimizing model parameters

exadel-inc / CompreFace

Leading free and open-source face recognition system

https://exadel.com/accelerator-showcase/compreface/

Apache License 2.0

5.63k stars 769 forks source link

how help optimizing model parameters #1069

Closed neozhu closed 1 year ago

neozhu commented 1 year ago

I am currently testing a model and have noticed a significant difference between two photos when compared. However, the similarity score between the two photos is still around 98%. I am using the latest version of docker-compose to deploy the testing environment. Can someone please advise me on how to adjust the model parameters to optimize its performance? Thank you in advance for your help!

pospielov commented 1 year ago

https://github.com/exadel-inc/CompreFace/blob/master/docs/Custom-builds.md#how-to-run-custom-builds I can recommend you run Mobilenet or SubCenter-ArcFace-r100 custom builds. They show better performance on Asians.

neozhu commented 1 year ago

Thank you very much for your assistance. Yes, switching to SubCenter-ArcFace-r100 has yielded excellent results.