ComfyUI Node
Magos DWP Extractor
A ComfyUI node in MAGOS Nodes/Temporal Editor with 14 inputs and 6 outputs.
Magos DWP Extractor
- images
- keyframe_data
- pose_data
- face_images
- bboxes
- facebboxes
- nlf_model
◄vitpose_model▾►
◄yolo_model▾►
◄nlf_model(None)►
◄onnx_deviceCUDAExecutionProvider►
◄detect_handstrue►
◄detect_facetrue►
◄detect_headtrue►
◄confidence_threshold0.30►
◄person_index0►
◄output_width0►
◄output_height0►
◄face_padding20►
◄debug_logfalse►
CategoryMAGOS Nodes/Temporal Editor
Inputs (14)
| Name | Type | Default | Description |
|---|---|---|---|
| images | IMAGE | — | |
| vitpose_model | COMBO | ViTPose ONNX model from ComfyUI/models/detection | |
| yolo_model | COMBO | YOLO ONNX model from ComfyUI/models/detection | |
| nlf_model | COMBO | (None) | NLF .safetensors from ComfyUI/models/nlf/ for 3D depth. Requires ComfyUI-SCAIL-Pose. Select (None) to skip. |
| onnx_device | COMBO | CUDAExecutionProvider | Device to run the ONNX models on |
| detect_hands | BOOLEAN | true | — |
| detect_face | BOOLEAN | true | — |
| detect_head | BOOLEAN | true | — |
| confidence_threshold | FLOAT | 0.300–1 | — |
| person_index | INT | 00–9 | — |
| output_width | INT | 00–8192 | 0 = use source width |
| output_height | INT | 00–8192 | 0 = use source height |
| face_padding | INT | 200–200 | — |
| debug_log | BOOLEAN | false | Write full trace to CMD + logs/session_*.log |
Outputs (6)
| Name | Type | Description |
|---|---|---|
| keyframe_data | KEYFRAME_DATA | — |
| pose_data | POSEDATA | — |
| face_images | IMAGE | — |
| bboxes | BBOX | — |
| facebboxes | BBOX | — |
| nlf_model | NLF_MODEL | — |