# NVIDIA-ISAAC-ROS/isaac_ros_visual_slam

Visual SLAM/odometry package based on NVIDIA-accelerated cuVSLAM

Repository: https://github.com/NVIDIA-ISAAC-ROS/isaac_ros_visual_slam
Canonical: https://ross.abutalabs.com/products/isaac_ros_visual_slam
Homepage: https://developer.nvidia.com/isaac-ros-gems
Language: C++
License: Apache-2.0
License Family: permissive
Topics: ros, visual-odometry, perception, slam, robotics, localization, ros2-humble, ros2, gpu, jetson
Last push: 2026-08-19T02:50:44+00:00

## Health v2 (maintenance only)
Score: 94/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 98, release rhythm 86, longevity 100
- inputs: {"age_days": 1785, "days_push": 14, "days_rel": 14, "gap_med": 43, "n_releases_24m": 12}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1448, forks 205 (observed 2026-08-28T04:04:45.526068+00:00)

## What it is
Isaac ROS Visual SLAM is a ROS 2 package providing GPU-accelerated visual simultaneous localization and mapping (VSLAM) using stereo visual inertial odometry (SVIO), built on NVIDIA's cuVSLAM. It estimates real-time, low-latency robot odometry from stereo cameras and optionally an IMU, serving as an input to navigation on mobile robots and drones.

## Use cases
- estimate robot odometry from stereo cameras in GPS-denied indoor environments
- provide localization for autonomous mobile robot navigation
- run real-time visual SLAM on NVIDIA Jetson embedded hardware
- fuse IMU data with stereo vision for robust visual-inertial odometry
- serve as the primary odometry source for drones
- integrate GPU-accelerated SLAM into an existing ROS 2 robotics stack

## When to choose
- you need real-time, low-latency visual odometry with GPU acceleration on NVIDIA hardware
- your robot operates indoors or where GPS is unavailable or unreliable
- you are building a ROS 2-based robotics application with stereo cameras and optionally an IMU
- you deploy on NVIDIA Jetson embedded platforms and want optimized performance

## When to avoid
- you do not have NVIDIA GPU or Jetson hardware, since acceleration depends on CUDA
- you need LiDAR-based SLAM rather than camera-based visual SLAM
- your environment lacks visual features and you have no IMU
- you are not using ROS 2 and cannot integrate with its middleware

## Facets
- artifact type: library
- maturity: active
- function: computer-vision, machine-learning, sdk
- domain: robotics, computer-vision, autonomous-vehicles
- platform: cpp
- tags: visual-slam, visual-odometry, stereo-camera, imu, localization, ros2, nvidia, gpu-accelerated, navigation, perception, linux, jetson, gpu

## Member repositories
- NVIDIA-ISAAC-ROS/isaac_ros_visual_slam (main) score 94

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:45.526068+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T04:36:02.765217+00:00, confidence not recorded.
  - readme: https://github.com/NVIDIA-ISAAC-ROS/isaac_ros_visual_slam (fetched 2026-08-28T04:04:45.526068+00:00, sha 9a3cd3b70802)
  - homepage: https://developer.nvidia.com/isaac-ros-gems (fetched 2026-08-29T11:45:40.836674+00:00, sha 3c484faed84f)
- Data as of 2026-08-30T08:39:29.467469+00:00.