diff --git a/README.md b/README.md index 3ceacef..b985c60 100644 --- a/README.md +++ b/README.md @@ -11,6 +11,8 @@ For all main contributors, please check [contributing](#contributing). ## Top News +**`2021-07-09`**: We add a [person_detection](examples/person_detection) example, trained by [SCRFD](detection/scrfd), which can be called directly by our [python-library](python-package). + **`2021-06-05`**: We launch a [Masked Face Recognition Challenge & Workshop](challenges/iccv21-mfr) on ICCV 2021. **`2021-05-15`**: We released an efficient high accuracy face detection approach called [SCRFD](detection/scrfd). diff --git a/examples/person_detection/README.md b/examples/person_detection/README.md index d3dfe86..e8fa6e2 100644 --- a/examples/person_detection/README.md +++ b/examples/person_detection/README.md @@ -1 +1,38 @@ -# Person Detection Powered by SCRFD +# Person Detection + +This person detection example is built by [SCRFD](../../detection/scrfd) approch. + +## Usage + +Firstly install insightface python library: + +``` +pip install -U insightface +``` + +and then load our person detection model by: + +``` +detector = insightface.model_zoo.get_model('scrfd_person_2.5g.onnx', download=True) +detector.prepare(0, nms_thresh=0.5, input_size=(640, 640)) +``` + +the model will be auto-downloaded from our storage server. + +## Detection Result: + +In this example, we support full-body detection and recognize the corresponding visible region in a single forward pass. + +Please see [scrfd_person.py](scrfd_person.py) for detail on how to visualize the results. + +The green bounding box shows the full-body while the blue mask indicates the visible region. + +(We make tests on the input size of 640) + + + + + + + +