GPU was not used for live demo

[Ishihara-Masabumi]

I tried to run some live demo, but I found that the execution was very slow and GPU was not used.
My environment is as follows:

Ubuntu20.04
Python3.8
tensorflow2.12.0
NVIDIA-SMI 530.30.02  Driver Version: 530.30.02    CUDA Version: 12.1

Please check this environment is OK or not, for live demo with GPU.

[AmmarkoV]

I think that the CUDA version is very new and the libtensorflow provided does not properly utilize it

Taking a look on the tables here :
https://www.tensorflow.org/install/source

For CUDA 12.2 you want tensorflow-2.15.0 Built with Clang 16.0.0, Bazel 6.1.0 and CUDNN 8.8

Try downloading
https://storage.googleapis.com/tensorflow/libtensorflow/libtensorflow-gpu-linux-x86_64-2.15.0.tar.gz

and extracting it in the dependencies/ directory!

You can also change
from
TENSORFLOW_VERSION="2.3.1"
to
TENSORFLOW_VERSION="2.15.0"
in
https://github.com/FORTH-ModelBasedTracker/MocapNET/blob/master/initialize.sh#L209
and then run the initialize.sh script selecting Y when asked for GPU