search

HamadYA / GhostFaceNets

This repository contains the official implementation of GhostFaceNets, State-Of-The-Art lightweight face recognition models.
https://ieeexplore.ieee.org/document/10098610
MIT License
194 stars 38 forks source link

Pre-trained model for face recognition #42

Closed arad2022 closed 6 months ago

arad2022 commented 8 months ago

How we can use the pre-trained model for face recognition? I have two person from each one 3 images and I want to extract embedding and compare them. Could you please provide any code example and he best way to compare embeddings?

@HamadYA

HamadYA commented 8 months ago

How we can use the pre-trained model for face recognition? I have two person from each one 3 images and I want to extract embedding and compare them. Could you please provide any code example and he best way to compare embeddings?

@HamadYA

Hi,

Please refer to this similar issue: https://github.com/leondgarse/Keras_insightface/issues/128

kiashann commented 8 months ago

How we can use the pre-trained model for face recognition? I have two person from each one 3 images and I want to extract embedding and compare them. Could you please provide any code example and he best way to compare embeddings? @HamadYA

Hi,

Please refer to this similar issue: leondgarse/Keras_insightface#128

Thanks for your answer. Could you please share checkpoints for GhostFaceNetV2? I just found weights for GhostFaceNetV1 in the repository. Which model has the high speed for real-time face recognition with good performance?

HamadYA commented 7 months ago

How we can use the pre-trained model for face recognition? I have two person from each one 3 images and I want to extract embedding and compare them. Could you please provide any code example and he best way to compare embeddings? @HamadYA

Hi, Please refer to this similar issue: leondgarse/Keras_insightface#128

Thanks for your answer. Could you please share checkpoints for GhostFaceNetV2? I just found weights for GhostFaceNetV1 in the repository. Which model has the high speed for real-time face recognition with good performance?

Hi,

Unfortunately GhostFaceNetV2 checkpoints are not available.

Models with small width multiplier are the real-time ones with good performance. All results with computational complexity you may find in the paper.