facedetector

This element can parse per-buffer inference tensor meta data generated by an upstream inference element.

Example launch command:

Test image files can be found here : https://github.com/Linzaer/Ultra-Light-Fast-Generic-Face-Detector-1MB/tree/master/imgs

The Model file can be found here : https://github.com/Linzaer/Ultra-Light-Fast-Generic-Face-Detector-1MB/tree/master/models/onnx

GST_DEBUG=facedetector
gst-launch-1.0 multifilesrc location=~/imgs/11.jpg ! jpegdec ! videoconvertscale !
onnxinference model-file=version-RFB-320.onnx input-image-format=chw input-tensor-offset=-127 input-tensor-scale=128.0 !
facedetector ! objectdetectionoverlay object-detection-outline-color=0xFF0000FF draw-labels=false !
videoconvertscale ! autovideosink

Hierarchy

GObject
    ╰──GInitiallyUnowned
        ╰──GstObject
            ╰──GstElement
                ╰──GstBaseTransform
                    ╰──facedetector

Factory details

Authors: – The original repository of the Ultra Light Face Detection is located at https://github.com/Linzaer/Ultra-Light-Fast-Generic-Face-Detector-1MB.Raghavendra Rao

Classification: – Tensordecoder/Video

Rank – primary

Plugin – tensordecoders

Package – GStreamer Bad Plug-ins

Pad Templates

`sink`

video/x-raw:

Presence – always

Direction – sink

Object type – GstPad

`src`

video/x-raw:

Presence – always

Direction – src

Object type – GstPad

Properties

iou-threshold

“iou-threshold” gfloat

Threshold for removing boxes based on proportion of the image

Flags : Read / Write

Default value : 0.3

Since : 1.28

score-threshold

“score-threshold” gfloat

Threshold for deciding when to remove boxes based on score

Flags : Read / Write

Default value : 0.6

Since : 1.28

The results of the search are