ZF/insightface
1
0
Fork 0
mirror of https://github.com/deepinsight/insightface.git synced 2025-12-30 08:02:27 +00:00
Code Issues Packages Projects Releases Wiki Activity

Merge branch 'master' of https://github.com/deepinsight/insightface

Browse Source
This commit is contained in:
Jia Guo
2021-06-20 00:17:19 +08:00
parent 6ab6e21469 a10e5e42fb
commit 4ea2cceaf8
3 changed files with 13 additions and 10 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
README.md
View File

@@ -34,7 +34,9 @@ InsightFace efficiently implements a rich variety of state of the art algorithms
### ArcFace Video Demo
[![ArcFace Demo](https://insightface.ai/assets/img/github/facerecognitionfromvideo.PNG)](https://www.youtube.com/watch?v=y-D1tReryGA&t=81s)
[<img src=https://insightface.ai/assets/img/github/facerecognitionfromvideo.PNG width="760" />](https://www.youtube.com/watch?v=y-D1tReryGA&t=81s)
Please click the image to watch the Youtube video. For Bilibili users, click [here](https://www.bilibili.com/video/av38041494?from=search&seid=11501833604850032313).

10
challenges/iccv21-mfr/README.md
View File

@@ -36,7 +36,7 @@ In this challenge, we will evaluate the accuracy of following testsets:
We ensure that there's no overlap between these testsets and public available training datasets, as they are not collected from online celebrities.
Our test datasets mainly comes from [IFRT](https://github.com/deepinsight/insightface/tree/master/challenges/IFRT).
Our test datasets mainly comes from [IFRT](../ifrt).
### ``Mask test-set:``
@@ -113,7 +113,7 @@ Inference time was evaluated on Tesla V100 GPU, using onnxruntime-gpu==1.6.
* Sub-Track A: Use MS1M-V3 as training set, download: [ref-link](https://github.com/deepinsight/insightface/tree/master/challenges/iccv19-lfr), feature length must <= 512, and the inference time must <= 10ms on Tesla V100 GPU.
* Sub-Track B: Use Glint360K as training set, download: [ref-link](https://github.com/deepinsight/insightface/tree/master/recognition/partial_fc), feature length must <= 1024, and the inference time must <= 20ms on Tesla V100 GPU.
2. Training set and testing set are both aligned to 112x112, re-alignment is prohibited.
3. Mask data-augmentation is allowed, such as [this](https://github.com/deepinsight/insightface/tree/master/recognition/tools). The applied mask augmentation tool should be reproducible.
3. Mask data-augmentation is allowed, such as [this](../../recognition/_tools_). The applied mask augmentation tool should be reproducible.
4. External dataset and pretrained models are both prohibited.
5. Participants submit onnx model, then get scores by our online evaluation. Test images are invisible.
6. Matching score is measured by cosine similarity.
@@ -128,9 +128,9 @@ Inference time was evaluated on Tesla V100 GPU, using onnxruntime-gpu==1.6.
## Tutorial
1. ArcFace-PyTorch (with Partial-FC), [code](https://github.com/deepinsight/insightface/tree/master/recognition/arcface_torch), [tutorial-cn](https://github.com/deepinsight/insightface/blob/master/challenges/iccv21-mfr/tutorial_pytorch_cn.md)
2. OneFlow, [code](https://github.com/deepinsight/insightface/tree/master/recognition/oneflow_face)
3. MXNet, [code](https://github.com/deepinsight/insightface/tree/master/recognition/ArcFace)
1. ArcFace-PyTorch (with Partial-FC), [code](../../recognition/arcface_torch), [tutorial-cn](tutorial_pytorch_cn.md)
2. OneFlow, [code](../../recognition/oneflow_face)
3. MXNet, [code](../../recognition/arcface_mxnet)
## Submission Guide

9
recognition/_tools_/README.md
View File

@@ -11,9 +11,9 @@ We can use this tool to do data augmentation while training our face recognition
| Face Image | OP | Mask Image | Out |
| ------- | ------ | --------- | ----------- |
| <img src="https://github.com/deepinsight/insightface/blob/master/deploy/Tom_Hanks_54745.png" alt="face" height="112" /> | +F | <img src="https://github.com/nttstar/insightface-resources/blob/master/images/mask1.jpg" alt="mask" height="112" /> | <img src="https://github.com/nttstar/insightface-resources/blob/master/images/mask_out1.jpg?raw=true" alt="mask" height="112" /> |
| <img src="https://github.com/deepinsight/insightface/blob/master/deploy/Tom_Hanks_54745.png" alt="face" height="112" /> | +F | <img src="https://github.com/nttstar/insightface-resources/blob/master/images/black-mask.png" alt="mask" height="112" /> | <img src="https://github.com/nttstar/insightface-resources/blob/master/images/mask_out3.jpg?raw=true" alt="mask" height="112" /> |
| <img src="https://github.com/deepinsight/insightface/blob/master/deploy/Tom_Hanks_54745.png" alt="face" height="112" /> | +H | <img src="https://github.com/nttstar/insightface-resources/blob/master/images/mask2.jpg?raw=true" alt="mask" height="112" /> | <img src="https://github.com/nttstar/insightface-resources/blob/master/images/mask_out2h.jpg?raw=true" alt="mask" height="112" /> |
| <img src="https://github.com/deepinsight/insightface/blob/master/python-package/insightface/data/images/Tom_Hanks_54745.png" alt="face" height="112" /> | +F | <img src="https://github.com/nttstar/insightface-resources/blob/master/images/mask1.jpg" alt="mask" height="112" /> | <img src="https://github.com/nttstar/insightface-resources/blob/master/images/mask_out1.jpg?raw=true" alt="mask" height="112" /> |
| <img src="https://github.com/deepinsight/insightface/blob/master/python-package/insightface/data/images/Tom_Hanks_54745.png" alt="face" height="112" /> | +F | <img src="https://github.com/nttstar/insightface-resources/blob/master/images/black-mask.png" alt="mask" height="112" /> | <img src="https://github.com/nttstar/insightface-resources/blob/master/images/mask_out3.jpg?raw=true" alt="mask" height="112" /> |
| <img src="https://github.com/deepinsight/insightface/blob/master/python-package/insightface/data/images/Tom_Hanks_54745.png" alt="face" height="112" /> | +H | <img src="https://github.com/nttstar/insightface-resources/blob/master/images/mask2.jpg?raw=true" alt="mask" height="112" /> | <img src="https://github.com/nttstar/insightface-resources/blob/master/images/mask_out2h.jpg?raw=true" alt="mask" height="112" /> |
**F** means FULL while **H** means HALF.
@@ -50,7 +50,8 @@ tool.prepare(ctx_id=0, det_size=(128,128)) #use gpu
(2) load face and mask images
```
image = cv2.imread("../../deploy/Tom_Hanks_54745.png")
from insightface.data import get_image as ins_get_image
image = ins_get_image('Tom_Hanks_54745')
mask_image = "mask_blue"
```