deepinsight / deepinsight/insightface

Verify the effect of VPL method

Open
#1,801 0 comments 0 reactions 0 assignees View on GitHub
Dominant language
Python
Stars
29.7k
Forks
6.1k
PR merge metrics
No merged PRs in 30d

Description

To verify the effect of VPL , I had trained two models with arcface loss:
One model with VPL {'start_iters': 20000, 'allowed_delta': 80, 'lambda': 0.15, 'mode': 0, 'momentum': False} , and theaccuracy of it is LFW 99.85, CFP_FP 99.15, AGEDB_30 98.38, IJBC 96.66 (TAR@FAR=0.0001) 99.57(AUC of ROC in IJBC)

And another model without VPL {'start_iters': 20000, 'allowed_delta': 80, 'lambda': 0.15, 'mode': -1, 'momentum': False}, and the accuracy of it is LFW 99.85, CFP_FP 99.10, AGEDB_30 98.43, IJBC 96.79 (TAR@FAR=0.0001) 99.55(AUC of ROC in IJBC)

other settings are same;

The VPL method don't seem to be working...

Contributor guide

No contributing guide indexed for this repository

Assessment

This issue has not been assessed yet.

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.