Tutorials Of Object Detection Model Server

Start A Object Detection Server

It’s very quick to start a object detection server. Main code are showed below

Object Detection Server Code Snappit

The executable binary file was built in $PROJECT_ROOT/_bin/yolov5_detection_server.out Simply run

cd $PROJECT_ROOT/_bin
./yolov5_detection_server.out ../conf/server/object_detection/yolov5/yolov5_server_config.ini

When server successfully start on http:://localhost:8091 you’re supposed to see worker_nums workers were called up and occupied your GPU resources. By default 4 model workers will be created you may enlarge it if you have enough GPU memory.

You may switch yolov5 model eg. yolov5s yolov5m etc by modifying model configuration. You may find instruction at about_model_configuration.md#L20

Python Client Example

Local python client test is similiar with mobilenetv2 classification server you may read toturials_of_classfication_model_server.md for details.

To use test python client you may run

cd $PROJECT_ROOT/scripts
export PYTHONPATH=$PWD:$PYTHONPATH
python server/test_server.py --server yolov5 --mode single

Unique Tips For Object Detection Model Python Client

Object deteciton model’s output is set of bounding boxes. A single bounding box consist of location, class_id and confidence. Server’s response is a json like

resp = {
    'req_id': '',
    'code': 1,
    'msg': 'success',
    'data': [
        {
            'cls_id': 6,
            'score': 0.65,
            'points': [[tl_x, tl_y], [rb_x, rb_y]],
            'detail_infos': {}
        },
        {
            ...
        },
    ]
}

Unique Tips For Face Detection Model Python Client

Face deteciton model’s output is set of bboxes. A single bounding box consist of location, landmarks and confidence. Server’s response is a json like

resp = {
    'req_id': '',
    'code': 1,
    'msg': 'success',
    'data': [
        {
            'cls_id': 6,
            'score': 0.65,
            'box': [[tl_x, tl_y], [rb_x, rb_y]],
            'landmark': [[x1, y1], [x2, y2], [x3, y3], ...]
        },
        {
            ...
        },
    ]
}

Object Detection Model’s Visualization Result

Yolov5 Model

Yolov5 :rocket: is a family of object detection architectures and models pretrained on the COCO dataset, and represents Ultralytics open-source research into future vision AI methods, incorporating lessons learned and best practices evolved over thousands of hours of research and development.

Server's Input Image

Server's Output Image With Different Model

LibFcae Model

Libface is a remarkable open source library for CNN-based face detection in images designed by ShiqiYu. You may refer to https://github.com/ShiqiYu/libfacedetection for details.

Server's Input Image

Server's Output Image