NVIDIA-ISAAC-ROS / NVIDIA-ISAAC-ROS/isaac_ros_visual_slam
How to use D455 with IMU in cuVSLAM?
@jaiveersinghNV is already working on this.
Since May 4, 2026.
- Dominant language
- C++
- Stars
- 1.5k
- Forks
- 206
- PR merge metrics
- No merged PRs in 30d
Description
System setup
A clear and concise description of the system setup.
Hardware ASUSTeK COMPUTER INC. ROG Strix G834JZ_G834JZ
GPU/driver/cuda version, etc. GTX 4080 for laptop Driver Version: 580.126.09 CUDA Version: 13.0
OS Ubuntu 24.04
Camera setup
A clear and concise description of the camera setup.
Camera model Realsense D455
Calibrated or not
Camera integration (to the robot / frame, etc)
Use 'N/A' if you are not using a camera
Description
I follow the tutorial https://nvidia-isaac-ros.github.io/concepts/visual_slam/cuvslam/tutorial_realsense.html# ,
but the trajectory I obtained is significantly deviated from reality, I do not know if the tutorial is just for D435i or smothing went wrong.
To Reproduce
Steps to reproduce the behavior.
Just use a D455 attach to computer, and run ros2 launch isaac_ros_visual_slam isaac_ros_visual_slam_realsense.launch.py
Expected behavior
A clear and concise description of what you expected to happen.
I want the trajectory reflect reality, and can be more rubust compare with the situation without IMU.
What you have tried
A clear and concise description of what you have tried so far.
I try another tutorial https://nvidia-isaac-ros.github.io/repositories_and_packages/isaac_ros_visual_slam/isaac_ros_visual_slam/index.html
Additional information
Add any other information / links / docs / datasets / rosbags about the problem here.
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Assessment
This issue has not been assessed yet.