2 Days to a demo

When importing jetson.inference and jetson.utils, where exactly can i find the methods and classes i use, like:

  • jetson.inference.imageNet(Network, [])
  • jetson.utils.cudaFont

I want to understand what exactly happens in the imagenet-camera.py, but can’t find the code to that.
What kind of feature extraction is used, is it a Cnn, how does the data gets reshaped to fit in the model?
I could not find any information about that.

I hope somebody can help me with that

Thank you,
Steffen

Hi steffen.epp, the Python bindings are found under /python/bindings. These bindings in turn call the C++ implementations found under /c Also the Python API documentation can be found here: https://github.com/dusty-nv/jetson-inference#api-reference

Yes, they are all DNN’s - see this list of the models: https://github.com/dusty-nv/jetson-inference#pre-trained-models

The pre-processing is done in a CUDA kernel, where resizing, mean pixel subtraction/normalization, and NCHW format are applied. For example, with imageNet, here is the PreProcess() function (this automatically gets called when you process the image with imageNet): https://github.com/dusty-nv/jetson-inference/blob/8846dcb274ff812880ef7bf9d42eb1dd75c9e7b7/c/imageNet.cpp#L389

The pre-processing CUDA kernels are found in: https://github.com/dusty-nv/jetson-inference/blob/master/c/imageNet.cu