# MobileFaceNet_TF

**Repository Path**: mjhapp/MobileFaceNet_TF

## Basic Information

- **Project Name**: MobileFaceNet_TF
- **Description**: No description available
- **Primary Language**: Unknown
- **License**: Apache-2.0
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 1
- **Created**: 2019-11-01
- **Last Updated**: 2020-12-19

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

### MobileFaceNet_TF

Tensorflow implementation for MobileFaceNet.

## dependencies

- tensorflow >= r1.5
- opencv-python 3.x
- python 3.x
- scipy
- sklearn
- numpy
- mxnet
- pickle

## Prepare dataset

1. choose one of the following links to download dataset which is provide by insightface. (Special Recommend MS1M-refine-v2)
* [MS1M-refine-v2@BaiduDrive](https://pan.baidu.com/s/1S6LJZGdqcZRle1vlcMzHOQ), [MS1M-refine-v2@GoogleDrive](https://www.dropbox.com/s/wpx6tqjf0y5mf6r/faces_ms1m-refine-v2_112x112.zip?dl=0)
* [Refined-MS1M@BaiduDrive](https://pan.baidu.com/s/1nxmSCch), [Refined-MS1M@GoogleDrive](https://drive.google.com/file/d/1XRdCt3xOw7B3saw0xUSzLRub_HI4Jbk3/view)
* [VGGFace2@BaiduDrive](https://pan.baidu.com/s/1c3KeLzy), [VGGFace2@GoogleDrive](https://www.dropbox.com/s/m9pm1it7vsw3gj0/faces_vgg2_112x112.zip?dl=0)
* [Insightface Dataset Zoo](https://github.com/deepinsight/insightface/wiki/Dataset-Zoo)
2. move dataset to `${MobileFaceNet_TF_ROOT}/datasets`.
3. run `${MobileFaceNet_TF_ROOT}/utils/data_process.py`.

## pretrained model

* [pretrained_model](https://github.com/sirius-ai/MobileFaceNet_TF/tree/master/arch/pretrained_model/)

## training

1. refined super parameters by yourself special project.
2. run script
`${MobileFaceNet_TF_ROOT}/train_nets.py`
3. have a snapshot result at `${MobileFaceNet_TF_ROOT}/output`.

## performance

|  size  | LFW(%) | Val@1e-3(%) | inference@MSM8976-cpu(ms) |
| ------ | ------ | ----------- | --------------------- |
|  5.7M  |  99.4+ |    98.4+    |          260-         |

## References

1. [facenet](https://github.com/davidsandberg/facenet)
2. [InsightFace mxnet](https://github.com/deepinsight/insightface)
3. [InsightFace_TF](https://github.com/auroua/InsightFace_TF)
4. [MobileFaceNets: Efficient CNNs for Accurate Real-Time Face Verification on Mobile Devices](https://arxiv.org/abs/1804.07573)
5. [CosFace: Large Margin Cosine Loss for Deep Face Recognition](https://arxiv.org/abs/1801.09414)
6. [InsightFace : Additive Angular Margin Loss for Deep Face Recognition](https://arxiv.org/abs/1801.07698)