From Friday, April 19th (11:00 PM CDT) through Saturday, April 20th (2:00 PM CDT), 2024, ni.com will undergo system upgrades that may result in temporary service interruption.

We appreciate your patience as we improve our online experience.

Machine Vision

cancel
Showing results for 
Search instead for 
Did you mean: 

Can not load .pb model into Labview 2018

I am trying to load my model which is trained by SVM and using HOG to extract feature. I also used torch to save this model after training (.pb file). But I cannot load .pb file into labview. Here is my .pb file and images

 

 

 

1.PNG

0 Kudos
Message 1 of 10
(4,924 Views)

Which version of tensorflow do you use ? 

 

ricky10116_real_0-1584694069397.png

0 Kudos
Message 2 of 10
(4,887 Views)

Please help me to solve this problem.Because I want to load this model (.pb file) to test my project. How can I do to load .pb file?

0 Kudos
Message 3 of 10
(4,880 Views)

Do you know your tensorflow version ? You need to use suggested version, other version is not working.  Because I make self-training model .pb in another version in the past, my solution is run code in other place. Not expert in the DL vi.....

0 Kudos
Message 4 of 10
(4,875 Views)

.pb file is a training character file which I used SVM algorithm in scikit-learn library. So how can I load .pb file into LabView to predict character ? Thank you so much 

0 Kudos
Message 5 of 10
(4,821 Views)

Should be like this

zdz_0-1622714013006.png

 

0 Kudos
Message 6 of 10
(4,204 Views)

 

 

I had this problem while running the model, please can someone help me ?

 

Capture.PNG

0 Kudos
Message 7 of 10
(3,902 Views)
Does the path of your pb model contain special characters?
0 Kudos
Message 8 of 10
(3,892 Views)

Did someone solved this issue?

 

I have the same problem with exactly the same error : 

LoisLbs_0-1671526013154.png

I created my model with Google AutoML and exported in .pb file.

0 Kudos
Message 9 of 10
(1,764 Views)

Clearly the issue is due to incompatibility with tensorflow version used in tensorflow LabVIEW API (TF1) and version used by Google AutoML to generate the model (TF2).

I think the only solution for now is to use python code to call model and then create a LabVIEW wrapper to call python code? 

0 Kudos
Message 10 of 10
(1,606 Views)