Detector

Usage

Python

Perform inference effortlessly by calling predict(...) as demonstrated below:

import json
from predict_shot import predict

with open("coco_converted/val/images/000056.jpg", 'rb') as file:
    image = file.read()
    results = predict(image)
print(json.dumps(results, indent=4))

Sample output:

[
    {
        "bbox_xyxy": [
            214.08851623535156,
            846.3001098632812,
            541.779296875,
            1198.1767578125
        ],
        "conf": 0.896393895149231
    },
    ...
]

Command Line Interface (CLI)

Run inference directly from the command line:

python predict_shot.py coco_converted/val/images/000056.jpg

Setting Up the Docker Container

If you need to launch your own Docker container, simply execute the following command:

bash run_docker.sh

For Machine Learning Engineers

Streamline your workflow with the full pipeline:

bash run_docker.sh
bash all_pipeline.sh

Elevate your ML tasks with these simple, yet powerful commands!