koide3 / glim

GLIM: versatile and extensible range-based 3D localization and mapping framework

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Slow processing speed

changh95 opened this issue · comments

Describe the bug

I ran GLIM on ROS1 Noetic docker image (koide3/glim_ros1:noetic_cuda12.2), with the provided downsized rosbag (os1_128_01_downsampled.bag).

It seems a lot slower than expected. I expected faster processing speed, from this video (https://www.youtube.com/watch?v=CIfRqeV0irE).

Is this processing speed normal?

Here's a video capture of my run.
https://youtu.be/q0Pk7DriU8s

To Reproduce

  1. Pull docker image from koide3/glim_ros1:noetic_cuda12.2.
  2. Start the docker container with docker run -it --rm --net=host --ipc=host --pid=host --gpus all -e=DISPLAY -e=ROS_DOMAIN_ID -v ./config:/glim/config -v /data:/data koide3/glim_ros1:noetic_cuda12.2
  3. roscore + rosrun glim_ros glim_rosnode + rosbag play /dta/os1_128_01_downsampled.bag

Expected behavior
I expected faster processing speed, like https://www.youtube.com/watch?v=CIfRqeV0irE

Screenshots
https://youtu.be/q0Pk7DriU8s

Environment (please complete the following information):

  • CPU/GPU: Core i9-13900K, RTX3080
  • OS: Ubuntu 20.04
  • ROS: ROS1 noetic docker
  • CUDA: CUDA 12.2
  • Build with CUDA: ON

To process faster, use glim_rosbag instead of glim_rosnode.

To process faster, use glim_rosbag instead of glim_rosnode.

Fixed!

glim_rosbag is super fast!! 👍 👍 👍

Closing issue as solved.