Eye close open clarification

General Discussion
1

Hi Sir,

    I would like to use Hailo8 with DeGirum tool to do face-recognization and then do the eye-open-close clarification. I got the onnx file from "https://huggingface.co/MichalMlodawski/open-closed-eye-classification-mobilev2" and convert it to .hef file under Hailo DFC environment. However, I already face the exception as below even I force the data input type to unit before I do the inference. Can you help?


Exception has occurred: DegirumException
Failed to perform model 'fan68_heatmap' inference: Model 'fan68_heatmap' inference failed: [ERROR]Incorrect value of parameter
Cannot reshape the tensor: incompatible linear sizes. Original linear size of shape x3 of type DG_INT8 is 3, while the linear size after reshaping to shape x1x1x1x1 of type DG_UINT8 is 1
dg_tensor_structs.h: 511 [DG::BasicTensor::reshapeTo]
When running model 'fan68_heatmap'
  File "/home/chi/h_test/hailo_examples/practice2.py", line 229, in process_fan68
    res = fan68_model(input_tensor)
          ^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/chi/h_test/hailo_examples/practice2.py", line 363, in <module>
    fan_res = process_fan68(fan68_model, crop)
degirum.exceptions.DegirumException: Failed to perform model 'fan68_heatmap' inference: Model 'fan68_heatmap' inference failed: [ERROR]Incorrect value of parameter
Cannot reshape the tensor: incompatible linear sizes. Original linear size of shape x3 of type DG_INT8 is 3, while the linear size after reshaping to shape x1x1x1x1 of type DG_UINT8 is 1
dg_tensor_structs.h: 511 [DG::BasicTensor::reshapeTo]
When running model 'fan68_heatmap'

Thanks,

BR

Ricky