github/deepface
1
1
Fork 0
mirror of https://github.com/serengil/deepface.git synced 2025-06-07 20:15:21 +00:00
Code Issues Packages Projects Releases Wiki Activity

best model and metric pair mentioned

Browse Source
This commit is contained in:
Sefik Ilkin Serengil 2020-07-05 17:04:49 +03:00 committed by GitHub
parent c912a3f063
commit e59c844b1b
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 8 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

12
README.md
View File

@ -16,7 +16,7 @@ pip install deepface
## Face Recognition ## Face Recognition
A modern [face recognition pipeline](https://sefiks.com/2020/05/01/a-gentle-introduction-to-face-recognition-in-deep-learning/) consists of 4 common stages: [detect](https://sefiks.com/2020/02/23/face-alignment-for-face-recognition-in-python-within-opencv/), [align](https://sefiks.com/2020/02/23/face-alignment-for-face-recognition-in-python-within-opencv/), [represent](https://sefiks.com/2018/08/06/deep-face-recognition-with-keras/) and [verify](https://sefiks.com/2020/05/22/fine-tuning-the-threshold-in-face-recognition/). **DeepFace** handles all these common stages in the background. A modern [**face recognition pipeline**](https://sefiks.com/2020/05/01/a-gentle-introduction-to-face-recognition-in-deep-learning/) consists of 4 common stages: [detect](https://sefiks.com/2020/02/23/face-alignment-for-face-recognition-in-python-within-opencv/), [align](https://sefiks.com/2020/02/23/face-alignment-for-face-recognition-in-python-within-opencv/), [represent](https://sefiks.com/2018/08/06/deep-face-recognition-with-keras/) and [verify](https://sefiks.com/2020/05/22/fine-tuning-the-threshold-in-face-recognition/). **DeepFace** handles all these common stages in the background.
**Face Verification** - [`Demo`](https://youtu.be/KRCvkNCOphE) **Face Verification** - [`Demo`](https://youtu.be/KRCvkNCOphE)
@ -44,9 +44,9 @@ df = DeepFace.find(img_path = "img1.jpg", db_path = "C:/workspace/my_db")
<p align="center"><img src="https://raw.githubusercontent.com/serengil/deepface/master/icon/stock-6-v2.jpg" width="95%" height="95%"></p> <p align="center"><img src="https://raw.githubusercontent.com/serengil/deepface/master/icon/stock-6-v2.jpg" width="95%" height="95%"></p>
**Supported face recognition models** **Face recognition models**
Deepface is a hybrid face recognition package. It currently wraps the **state-of-the-art** face recognition models: [`VGG-Face`](https://sefiks.com/2018/08/06/deep-face-recognition-with-keras/) , [`Google FaceNet`](https://sefiks.com/2018/09/03/face-recognition-with-facenet-in-keras/), [`OpenFace`](https://sefiks.com/2019/07/21/face-recognition-with-openface-in-keras/), [`Facebook DeepFace`](https://sefiks.com/2020/02/17/face-recognition-with-facebook-deepface-in-keras/) and [`DeepID`](https://sefiks.com/2020/06/16/face-recognition-with-deepid-in-keras/). The default configuration verifies faces with **VGG-Face** model. You can set the base model while verification as illustared below. Accuracy and speed show difference based on the performing model. Deepface is a hybrid face recognition package. It currently wraps the **state-of-the-art** face recognition models: [`VGG-Face`](https://sefiks.com/2018/08/06/deep-face-recognition-with-keras/) , [`Google FaceNet`](https://sefiks.com/2018/09/03/face-recognition-with-facenet-in-keras/), [`OpenFace`](https://sefiks.com/2019/07/21/face-recognition-with-openface-in-keras/), [`Facebook DeepFace`](https://sefiks.com/2020/02/17/face-recognition-with-facebook-deepface-in-keras/) and [`DeepID`](https://sefiks.com/2020/06/16/face-recognition-with-deepid-in-keras/). The default configuration verifies faces with **VGG-Face** model. You can set the base model while verification as illustared below.
```python ```python
models = ["VGG-Face", "Facenet", "OpenFace", "DeepFace", "DeepID"] models = ["VGG-Face", "Facenet", "OpenFace", "DeepFace", "DeepID"]
@ -54,6 +54,8 @@ for model in models:
result = DeepFace.verify("img1.jpg", "img2.jpg", model_name = model) result = DeepFace.verify("img1.jpg", "img2.jpg", model_name = model)
``` ```
FaceNet and VGG-Face [overperforms](https://youtu.be/i_MOwvhbLdI) than OpenFace, DeepFace and DeepID based on experiments.
**Similarity** **Similarity**
Face recognition models are regular [convolutional neural networks](https://sefiks.com/2018/03/23/convolutional-autoencoder-clustering-images-with-neural-networks/) and they are responsible to represent face photos as vectors. Decision of verification is based on the distance between vectors. We can classify pairs if its distance is less than a [threshold](https://sefiks.com/2020/05/22/fine-tuning-the-threshold-in-face-recognition/). Face recognition models are regular [convolutional neural networks](https://sefiks.com/2018/03/23/convolutional-autoencoder-clustering-images-with-neural-networks/) and they are responsible to represent face photos as vectors. Decision of verification is based on the distance between vectors. We can classify pairs if its distance is less than a [threshold](https://sefiks.com/2020/05/22/fine-tuning-the-threshold-in-face-recognition/).
@ -66,6 +68,8 @@ for metric in metrics:
result = DeepFace.verify("img1.jpg", "img2.jpg", distance_metric = metric) result = DeepFace.verify("img1.jpg", "img2.jpg", distance_metric = metric)
``` ```
Euclidean L2 form [overperforms](https://youtu.be/i_MOwvhbLdI) than cosine and regular Euclidean distance based on experiments.
**Ensemble learning for face recognition** - [`Demo`](https://youtu.be/EIBJJJ0ECXU) **Ensemble learning for face recognition** - [`Demo`](https://youtu.be/EIBJJJ0ECXU)
A face recognition task can be handled by several models and similarity metrics. Herein, deepface offers a [special boosting and combination solution](https://sefiks.com/2020/06/03/mastering-face-recognition-with-ensemble-learning/) to improve the accuracy of a face recognition task. This provides a huge improvement on accuracy metrics. Human beings could have 97.53% score for face recognition tasks whereas this ensemble method passes the human level accuracy and gets 98.57% accuracy. On the other hand, this runs much slower than single models. A face recognition task can be handled by several models and similarity metrics. Herein, deepface offers a [special boosting and combination solution](https://sefiks.com/2020/06/03/mastering-face-recognition-with-ensemble-learning/) to improve the accuracy of a face recognition task. This provides a huge improvement on accuracy metrics. Human beings could have 97.53% score for face recognition tasks whereas this ensemble method passes the human level accuracy and gets 98.57% accuracy. On the other hand, this runs much slower than single models.
@ -153,7 +157,7 @@ DeepFace.analyze("img1.jpg", models=models)
## E-Learning ## E-Learning
Deepface package for python is mentioned in this [playlist](https://www.youtube.com/watch?v=KRCvkNCOphE&list=PLsS_1RYmYQQFdWqxQggXHynP1rqaYXv_E) as video lectures. Subscribe the channel to stay up-to-date and be informed when a new lecture is added. Deepface package for python is mentioned in this [playlist](https://www.youtube.com/watch?v=KRCvkNCOphE&list=PLsS_1RYmYQQFdWqxQggXHynP1rqaYXv_E) as video lectures. **Subscribe** the channel to stay up-to-date and be informed when a new lecture is added.
## Translations ## Translations