M8 to onnx

[HandsLing]

can you tell me how to tranfer M8 model to onnx? thanks

[pourfard]

I trained and uploaded onnx model.

[Kk875]

I trained and uploaded onnx model.

hi, I want to know how to train M8 model and how to prepare the dataset? could you tell me? thank you

[HandsLing]

Hello, I saw you mentioned that the accuracy of combining models m6 and m8 exceeds 99%, but I noticed that the example code you provided is for individual inference. Could you please provide the code for combining these models? Thank you.

[pourfard]

Hi,

Here is a sample code:

real_scores = []

for liveness_detector in liveness_detectors:
    real_scores.append(liveness_detector.get_real_score(bgr, bbox[:4]))

if len(real_scores) == 1:
    real_score = real_scores[0]
else:
    real_score = min(real_scores)

if real_score < MIN_REAL_SCORE:
    return ERR_SPOOF_DETECTION

If you want to use several models, you have several options:
1- Pessimistic: min(real_scores). I use this option.
2- Optimistic: max(real_scores) or avg(real_scores)

[pourfard]

could

I have just trained a simple binary classifier (mobilenet-v3) . You can read the docs.