Profile My skills Saved Collections Edit profile Submit a skill

← all publishers

nvidia-isaac

@nvidia-isaac source repo

5 published skills

  1. ▌
    Setup · nvidia-isaac bundle
    Installs and verifies the perception pipeline environment — the pipeline's own venv, SAM3, the FoundationPose Inference Library, and FoundationStereo run through TAO Deploy as a TensorRT engine. Use when asked to set up, install, provision, bootstrap, or repair this repo's environment, or when a run fails with a missing-dependency, venv, CUDA, TensorRT, pycuda, or LD_LIBRARY_PATH error.
    0
    installs
  2. ▌
    Pipeline · nvidia-isaac bundle
    Runs one dataset end to end (depth → SAM3 → FoundationPose) on the COMMERCIAL depth model — the TAO Deploy TensorRT engine — and reads out the pose metrics, including the flags a dataset without collected depth needs and how to prove which model actually ran. Covers adapting a BOP dataset into the pipeline's layout first, which the engine build depends on. Use when asked to run, evaluate, benchmark, or re-run a dataset, to adapt or convert a BOP dataset, to regenerate depth for one, or to compare pose results before and after a change. Environment installation is out of scope — that is the "setup" skill.
    0
    installs
  3. ▌
    Cuvslam CI · nvidia-isaac bundle
    Use when working on cuVSLAM CI/CD - the GitHub Actions nightly and PR-verify pipelines, dataset provisioning and staging, the evaluation run and KPI reporting, the build/test/lint matrix, branch rulesets, or repository secrets and variables. Covers adding a dataset, changing dataset format or packing, and controlling which datasets run in PR versus nightly.
    0
    installs
  4. ▌
    Cuvslam Onboard · nvidia-isaac bundle
    Build, install, and run NVIDIA cuVSLAM and PyCuVSLAM from source or wheels. Covers environment setup, dataset preparation, and running examples for all tracking modes (stereo, mono, mono-depth, stereo-inertial, multi-camera, and Multisensor) and SLAM (mapping, localization, loop closure). Use when asked to: build cuVSLAM, install PyCuVSLAM, set up cuVSLAM environment, run cuVSLAM examples, prepare KITTI/EuRoC/TUM datasets, run visual odometry, set up live camera tracking (RealSense/ZED/OAK-D/Orbbec), run cuVSLAM in Docker, or use cuVSLAM C++ tools.
    0
    installs
  5. ▌
    Cuvslam Troubleshoot · nvidia-isaac bundle
    Troubleshoot NVIDIA cuVSLAM (CUDA Visual SLAM) pose accuracy, tracking, build, and integration issues. Use when asked to debug cuVSLAM tracking failures, pose drift, lost tracking, calibration problems, image quality issues, IMU integration, multi-camera setups, SLAM loop closure, or build/install problems with cuVSLAM, PyCuVSLAM, or Isaac ROS cuVSLAM. Triggers on: "cuVSLAM not tracking", "visual odometry drift", "SLAM lost", "cuvslam build error", "PyCuVSLAM install", "Isaac ROS visual_slam", "camera calibration issue", "stereo tracking bad", "cuVSLAM pose inaccurate".
    0
    installs