Karan Chawla

Results 8 comments of Karan Chawla

This is meant to be used as a library. You can preprocess the data and run it in real time as you're getting your sensor measurements. Use the prediction loop...

Since I didn't find a solution to this, I just got Parallels and am using ROS with Ubuntu. It's just easier.

I used this on the Kitti dataset and this information was extracted from the ground truth provided in the dataset.

http://www8.cs.umu.se/kurser/5DV122/HT13/material/Hellstrom-ForwardKinematics.pdf You can find the equations here. I also used the same on my [rover](https://github.com/karanchawla/CallistoRover/blob/master/src/diffdrive_odom/src/diffdrive_odom.py).

https://goo.gl/photos/gzKaa8xhyTfy8dKv6 This link shows the aforementioned odometry working. Here I am using a simple closed loop PID control with feedback from the encoders on my left and right wheel. For...

Are you referring to training times or during prediction?