serengil/deepface
Dockerized DeepFace: A Containerized Face Recognition and Facial Attribute Analysis Service
100K+
serengil/deepface repository overview
Dockerized DeepFace: A Containerized Face Recognition and Facial Attribute Analysis Service
Repo: https://github.com/serengil/deepface
Video Tutorial: https://youtu.be/9Tk9lRQareA
Running The Service
Firstly, pull the latest deepface image
docker pull serengil/deepface
Secondly, run the deepface via docker
docker run -p 5005:5000 deepface
This will get the deepface service up at http://localhost:5005. Confirm it is accessible from the browser.
Now, you are able to call deepface functionalities.
This postman collection will guide you how to call deepface functionalities.
Face Verification
You can call the face verification method at the endpoint http://localhost:5005/verify using an HTTP POST request with the following body:
{
"img1_path": "https://raw.githubusercontent.com/serengil/deepface/master/tests/dataset/img1.jpg",
"img2_path": "https://raw.githubusercontent.com/serengil/deepface/master/tests/dataset/img2.jpg"
},
"model_name": "VGG-Face",
"detector_backend": "opencv",
"distance_metric": "cosine"
The response will look like this:
{
"detector_backend": "opencv",
"distance": 0.42272907319751785,
"facial_areas": {
"img1": {
"h": 768,
"left_eye": [
850,
524
],
"right_eye": [
571,
517
],
"w": 768,
"x": 339,
"y": 218
},
"img2": {
"h": 491,
"left_eye": [
858,
388
],
"right_eye": [
663,
390
],
"w": 491,
"x": 524,
"y": 201
}
},
"model": "VGG-Face",
"similarity_metric": "cosine",
"threshold": 0.68,
"time": 3.39,
"verified": true
}
Check the verified key in the response payload for verification status.
You can also send base64 encoded strings instead of URLs.
Facial Attribute Analysis
This method allows you to predict age, gender, emotion, and race/ethnicity. The endpoint address is http://127.0.0.1:5005/analyze. Use an HTTP POST request with the following body:
{
"img_path": "https://raw.githubusercontent.com/serengil/deepface/master/tests/dataset/img1.jpg",
"actions": ["age", "gender", "emotion", "race"]
}
The response will provide predictions on apparent age, gender, emotion, and ethnicity:
{
"results": [
{
"age": 31,
"dominant_emotion": "sad",
"dominant_gender": "Man",
"dominant_race": "white",
"emotion": {
"angry": 10.380517944035775,
"disgust": 0.34448695919073113,
"fear": 41.99173247912735,
"happy": 0.031012031045239443,
"neutral": 2.4377143829805394,
"sad": 44.47406298231328,
"surprise": 0.34046527596231985
},
"face_confidence": 0.89,
"gender": {
"Man": 99.99207258224487,
"Woman": 0.007924179953988642
},
"race": {
"asian": 1.6582146286964417,
"black": 0.2634486649185419,
"indian": 1.1727581731975079,
"latino hispanic": 41.77250564098358,
"middle eastern": 10.031893849372864,
"white": 45.10117769241333
},
"region": {
"h": 681,
"left_eye": null,
"right_eye": null,
"w": 681,
"x": 1436,
"y": 336
}
},
{
"age": 31,
"dominant_emotion": "neutral",
"dominant_gender": "Woman",
"dominant_race": "white",
"emotion": {
"angry": 0.05392062594182789,
"disgust": 0.00015850500858505256,
"fear": 0.35890184808522463,
"happy": 43.93383264541626,
"neutral": 54.20163869857788,
"sad": 1.254333183169365,
"surprise": 0.19721481949090958
},
"face_confidence": 0.92,
"gender": {
"Man": 0.021838860993739218,
"Woman": 99.97816681861877
},
"race": {
"asian": 0.25855733547359705,
"black": 0.016115940525196493,
"indian": 0.2287688199430704,
"latino hispanic": 5.218064412474632,
"middle eastern": 8.171910792589188,
"white": 86.10658645629883
},
"region": {
"h": 758,
"left_eye": [
833,
819
],
"right_eye": [
553,
860
],
"w": 758,
"x": 326,
"y": 542
}
}
]
}
Like face verification, this method also supports base64 encoded strings instead of URLs.
Represent
This method generates vector embeddings for given images. The endpoint address is http://127.0.0.1:5005/represent. Use an HTTP POST request with the following body:
{
"model_name": "VGG-Face",
"detector_backend": "opencv",
"img": "https://raw.githubusercontent.com/serengil/deepface/master/tests/dataset/img1.jpg"
}
The response will include the vector embeddings:
{
"results": [
{
"embedding": [
0.013881363057331663,
0.0865466348140943,
0.00843181578293164,
0.08922029773772754,
0.009442750920063645,
0.0585725425777175,
0.010859464741657977,
# ...
],
"face_confidence": 0.92,
"facial_area": {
"h": 768,
"left_eye": [
850,
524
],
"right_eye": [
571,
517
],
"w": 768,
"x": 339,
"y": 218
}
}
]
}
Citation
If you use deepface in your research, please cite this publication:
@article{serengil2026boosted,
title = {Boosted LightFace: A Hybrid DNN and GBM Model for Boosted Facial Recognition},
author = {Serengil, Sefik Ilkin and Ozpinar, Alper},
journal = {Gazi University Journal of Science},
volume = {39},
number = {1},
pages = {452-466},
year = {2026},
doi = {10.35378/gujs.1794891},
url = {https://dergipark.org.tr/en/pub/gujs/article/1794891},
publisher = {Gazi University}
}
Tag summary
Content type
Image
Digest
sha256:58d96a2ff…
Size
1.7 GB
Last updated
10 months ago
docker pull serengil/deepface