This demo uses Milvus Vector Database for storing face embeddings & performing face similaity search based on those stored embeddings. Milvus is a vector similarity search engine, a tool that lets you quickly find the closest matching vector in a pool of billions of vectors.
The system architecture is as below:
This demo uses the dataset of around 800k images consisting of 1100 Famous Celebrities and an Unknown class to classify unknown faces. All the images have been scraped from Google and contains no duplicate images. Each Celebrity class(folder) consists approximately 700-800 images and the Unknown class consists of 100k images.
- Download the following dataset(zip) inside
quick_deploy/server/src: Celeb Dataset - Optional but recommended: Download the encoded celebrity files(will save a lot of time due to slow processing of images): Encodings
Note: You can also use other images for testing. This system supports the following formats: .jpg and .png.
The face recognition bootcamp system requires Milvus. We can start these containers with one click through docker-compose.yaml.
- Modify docker-compose.yaml to map your data directory to the docker container
$ git clone https://github.com/Spnetic-5/bootcamp.git
$ cd solutions/image/face_recognition_system/quick_deploy- Create containers & start servers with docker-compose.yaml
$ docker-compose up -dThen you will see the that all containers are created.
[+] Running 12/12
⠿ standalone Pulled 28.5s
⠿ 171857c49d0f Pull complete 4.6s
⠿ 419640447d26 Pull complete 4.9s
⠿ 61e52f862619 Pull complete 5.0s
⠿ 2580b47486e5 Pull complete 20.3s
⠿ cd742921730d Pull complete 22.2s
⠿ 936cb7027fe4 Pull complete 22.2s
⠿ 319dd389c04d Pull complete 24.2s
⠿ 543c11caaeb6 Pull complete 24.3s
⠿ 06d62b89360c Pull complete 24.5s
⠿ 5186d5863148 Pull complete 24.6s
⠿ b410b80e82c0 Pull complete 24.7s
[+] Running 3/3
⠿ Container milvus-minio Started 1.3s
⠿ Container milvus-etcd Started 1.3s
⠿ Container milvus-standalone Started And show all containers with docker ps:
CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES
47952d6972fb milvusdb/milvus:v2.0.2 "/tini -- milvus run…" 38 minutes ago Up 38 minutes 0.0.0.0:19530->19530/tcp, :::19530->19530/tcp milvus-standalone
870d464532e1 quay.io/coreos/etcd:v3.5.0 "etcd -advertise-cli…" 38 minutes ago Up 38 minutes 2379-2380/tcp milvus-etcd
0b0219549307 minio/minio:RELEASE.2020-12-03T00-03-10Z "/usr/bin/docker-ent…" 38 minutes ago Up 38 minutes (healthy) 9000/tcp milvus-minioWe recommend using Docker Compose to deploy the face recognition bootcamp. However, you also can run from source code, you need to manually start Milvus and MySQL. Next show you how to run the API server and Client.
First, you need to start Milvus & MySQL servers.
Refer Milvus Standalone for how to install Milvus. Please note the Milvus version should match pymilvus version in config.py.
There are several ways to start Mysql. One option is using docker to create a container:
- Download Configuration Files
$ mkdir -p /home/$USER/milvus/conf
$ cd /home/$USER/milvus/conf
$ wget https://raw.githubusercontent.com/milvus-io/milvus/v1.1.0/core/conf/demo/server_config.yaml$ sudo docker run -d --name milvus_cpu_1.1.0 -p 19530:19530 -p 19121:19121 -v /home/$USER/milvus/db:/var/lib/milvus/db -v /home/$USER/milvus/conf:/var/lib/milvus/conf -v /home/$USER/milvus/logs:/var/lib/milvus/logs -v /home/$USER/milvus/wal:/var/lib/milvus/wal milvusdb/milvus:1.1.0-cpu-d050721-5e559cThen to start the system server, and it provides HTTP backend services.
- Install the Python packages
$ git clone https://github.com/Spnetic-5/bootcamp.git
$ cd solutions/face_recognition_system/quick_deploy/server
$ pip install -m requirements.txt- For Milvus v1.1.0
$ pip3 install pymilvus==1.1.0- For Milvus v2.1.0
$ pip3 install pymilvus==2.1.3- Set configuration
Modify the parameters according to your own environment. Here listing some parameters that need to be set, for more information please refer to config.py.
| Parameter | Description | Default setting |
|---|---|---|
| MILVUS_HOST | The IP address of Milvus, you can get it by ifconfig. | 127.0.0.1 |
| MILVUS_PORT | Port of Milvus. | 19530 |
| VECTOR_DIMENSION | Dimension of the vectors. | 1000 |
| MYSQL_HOST | The IP address of Mysql. | 127.0.0.1 |
| MYSQL_PORT | Port of Mysql. | 3306 |
| DEFAULT_TABLE | The milvus and mysql default collection name. | milvus_img_search |
Prepare the dataset for Milvus Search Engine
python3 src/prepare_data.py- Run the code
Then start the server.
- Replace test.jpg with <image_path>
python3 src/search_face.py test.jpgIf you are interested in our code or would like to contribute code, feel free to learn more about our code structure.
server
├── Dockerfile
├── requirements.txt
└── src
├── celeb_reorganized
├── volumes
├── config.py # Configuration file
├── encoded_save.npy # Convert an image to embedding (ResNet50)
├── example.py # Milvus starter example
├── search_face.py # Source code to start webserver
├── search_face_milvusv1.py # Source code to start webserver with Milvus v1.1.0
├── search_face_milvusv2.py # Source code to start webserver with Milvus v2.1.0
├── id_to_class
├── identity_CelebA.txt
├── identity_save.npy
├── prepare_data.py
└── test_main.py # Pytest file for main.py