• Community

  • Converting SSD MobileNet to Blob using online converter

I'm trying to convert a SSD MobileNet V2 FPNLite 320*320 frozen graph to a blob using the online converter. I get the following .. [ FRAMEWORK ERROR ] Cannot load input model: TensorFlow cannot read the model file: saved_model.pb is incorrect TensorFlow model file. Just wondering if I am doing something wrong or if this particular model requires in-depth modifications to be converted successfully?

Hey, I assume you are using a model with TF Object Detection API. For this, some additional parameters are required when exporting the model, though I do not know what the relative path with online blob converter is. CC @Luxonis-Lukasz

Meanwhile, I'd suggest you checkout this documentation from OpenVINO and try to convert the model locally. You can also refer to our tutorial, which shows an example of converting a similar model.

Best,
Matija

Thanks for the info, I will look further into local conversion. I have already trained two custom models using the colab notebook provided. They both perform well, but one of them is not quite as good as the SSD MobileNet V2 FPNLite 320*320 that I previously created. As it stands, my training images where taken with a different camera so I am adding images from my OAK-D and will retest. This by itself may be enough to improve the accuracy.