ReadNetFromTensorflow differences in python and c++

1

Hi,

at the moment I try to read in a self-built tensorflow model. Using python it is no problem to use the ReadNetFromTensorflow method without an existing config file. In C++ I tried different versions of the ReadNetFromTensorflow method and not a single is working without a config file. Even the cv::dnn::readNetFromTensorflow(const char* bufferModel ,size_t lenModel, const char* bufferConfig = NULL , size_t lenConfig = 0) method isn’t working without a config file.
Do I necessarily need a config file, and if so why can I use this method in python without any errors?

I’m thankfull for any help and support.

2

this is the wrong overload, it does NOT expect filenames, but memory buffers

please try this instead:
https://docs.opencv.org/4.x/d6/d0f/group__dnn.html#gad820b280978d06773234ba6841e77e8d