I’m having trouble properly configuring the compilation environment and running the training script.

Hello there! @angadbajwa

I came across your article about IEEE and summarized it as follows: a residual convolutional network model was designed to output a velocity observation using IMU data, while an extended Kalman filter uses acceleration, angular velocity, and other IMU data to produce a velocity prediction. The observation is then incorporated into a correction term and fused with the prediction to ultimately obtain a more accurate prediction.

I wanted to ask if you could provide more detailed information about this project. I’ve carefully reviewed your code, but I’m still confused. In the training script, I’m unable to install the spherical_coords_comparator and pymlg libraries. Is it because these libraries are not open-source? Or could it be an issue with the compatibility between my Python 3.7 and pip versions? Additionally, could you share the README and requirements.txt files? I’d really like to try out this impressive project.

Looking forward to your reply.