Detects faces in the input image and outputs boxes; enabling output_landmarks adds 478-point landmarks (mediapipe_478 scheme, unified across backends). With backend = auto, MediaPipe runs first and MNN takes over when unavailable; the backend actually used is observable in the result.
Module：Face detection · Task type：face_detect
Ports
Inputs
| Port | Type | Required |
|---|---|---|
in | Image / cv::Mat | ✓ |
Outputs
| Port | Type | Required |
|---|---|---|
out | Image / cv::Mat | — |
Parameters
| 参数 | 类型 | 默认值 | 范围 / 选项 | 说明 |
|---|---|---|---|---|
backend | enum | 0 | Auto / MediaPipe / MNN | inference backend; Auto degrades gracefully (MediaPipe → MNN) |
model_path | string | （空） | *shown conditionally on backend。*MNN detector model path (empty = built-in search path) | |
landmark_model_path | string | （空） | *shown conditionally on backend。*landmark model path (used with output_landmarks) | |
output_landmarks | bool | false | emit 478-point landmarks | |
max_faces | int | 5 | [1, 10] | maximum faces kept |
score_threshold | float | 0.5 | [0, 1] | confidence threshold |
nms_threshold | float | 0.3 | [0, 1] | NMS overlap threshold |
delegate | enum | 0 | CPU / GPU | *shown conditionally on backend。*MediaPipe inference delegate (CPU / GPU) |
device | enum | 2 | CPU / Metal / Auto | *shown conditionally on backend。*MNN device (CPU / Metal / Auto) |
threads | int | 4 | [1, 16] | *shown conditionally on backend。*CPU inference threads |
Example graph
A minimal runnable pipeline (read → process → write). Drag it into GraphStudio or open via --open; replace data/test.png with a real path.
{
"version": "2.0",
"tasks": [
{
"id": "read",
"type": "opencv_image_read",
"params": {
"file_path": "data/test.png"
}
},
{
"id": "node",
"type": "face_detect",
"params": {
"backend": 0,
"output_landmarks": false,
"max_faces": 5,
"score_threshold": 0.5,
"nms_threshold": 0.3,
"delegate": 0,
"device": 2,
"threads": 4
}
},
{
"id": "save",
"type": "opencv_image_write",
"params": {
"file_path": "out_out.png"
}
}
],
"edges": [
{
"from": "read",
"from_port": "out",
"to": "node",
"to_port": "in"
},
{
"from": "node",
"from_port": "out",
"to": "save",
"to_port": "in"
}
]
}
Notes
- Missing model assets fail the task with a pointer to the download script — never silent empty results.
- TG_FACE_DEBUG=1 enables backend diagnostics.
Part of the GraphStudio node guide series; parameters reflect the latest release.