how to using onnx model without cv FaceDetectorYN

[NinjaCats]

I didn't find relevant code in <opencv2/objdetect/face.hpp>. Could you please explain the meaning of these outputs? I want to integrate them into another platform where I can't use OpenCV

this is face_detection_yunet_2023mar.onnx outputs
0 cls_8: [1, 6400, 1]
1 cls_16: [1, 1600, 1]
2 cls_32: [1, 400, 1]
3 obj_8: [1, 6400, 1]
4 obj_16: [1, 1600, 1]
5 obj_32: [1, 400, 1]
6 bbox_8: [1, 6400, 4]
7 bbox_16: [1, 1600, 4]
8 bbox_32: [1, 400, 4]
9 kps_8: [1, 6400, 10]
10 kps_16: [1, 1600, 10]
11 kps_32: [1, 400, 10]

[darnellcoco]

Also meet this problem, do u get the meaning of these outputs now?