Wrong detected face saved

[Jaroslove]

Hi Team,

I use facenet_pytorch to detect faces in a image, save extracted faces to a blob storage and get vectors from extracted faces.

I wrote a following code:

        file_bytes: bytes = await file.read()
        nparr = np.frombuffer(file_bytes, np.uint8)
        image = cv2.imdecode(nparr, cv2.IMREAD_COLOR)
        cropped = mtcnn(image)
        if cropped is not None:
            name = f"{uuid.uuid4()}.jpg"
            full_path = os.path.join(image_blob_storage_path, name)
            cropped = cropped.numpy()
            if isinstance(cropped, np.ndarray):
                cv2.imwrite(full_path, cv2.cvtColor(cropped, cv2.COLOR_RGB2BGR))
            else:
                cropped.save(full_path)
            return encode(cropped)
        return []

When I test this code extracted faces saved but I cannot see any face on saved images, it's just dots.
Could you provide for me any help, please?