Use the NuRec Render Command#

To render views from a trained NuRec model, use the render command. The render command takes a USDZ file as the model input and doesn’t require the original input clip. You can use it instead of rendering through the NuRec gRPC API.

The render command supports rendering from a training trajectory, or from a trajectory that is at a fixed rotation/translation away from each training pose. It does not currently support user-fed trajectory and intrinsics. It is limited to the intrinsics of a training camera, and you must pass the camera name when you run the command.

For example, to render from the original training views, you would use the following command (with the proper USDZ file path and output directories set):

docker run --shm-size=64g -it --rm --gpus all \
    --network host \
    --volume /path/to/output/folder:/workdir/output \
    nvcr.io/nvidia/nre/nre-ga:26.04 \ 
    -- render \
    --artifact-path "${{USDZ_PATH}}" \
    --output-dir "${{RENDER_OUTPUT_DIR}}" \
    --frame-step 1 \
    --image-scale 0.25 \
    --camera-id camera_front_wide_120fov \
    --camera-id camera_front_tele_30fov \
    --image-format png \
    --frame-naming frame-end-timestamp \
    --replicate-training-views

The --replicate-training-views option ensures that the trained per-frame ISP parameters are applied, otherwise the pixel intensities and colors may differ from those in the training images. Use --no-replicate-training-views if you are planning to apply transformations, generating novel views, using the --rolling-shutter-duration override, or passing the --enable-editing-actors flag.

Apply fixed transformations in the vehicle space (see the conventions) to every pose of the rig trajectory by appending the following options.

  --rig-rotation-offset <yaw> <-roll> <pitch>  --rig-translation-offset <tx> <ty> <tz>

The --enable-editing-actors flag allows modifications to dynamic actor poses and only renders the dynamic actors present within the current frame’s start and end timestamp. The --demo-actor-transform flag applies a precomputed transformation (+1m Z-lift) to discriminate the editable (dynamic) actors by lifting them from the road surface in the rendered images. If you’re passing the --demo-actor-transform flag, you must also pass the --enable-editing-actors flag.

Use --render-bboxes to overlay 3D cuboid wireframes onto rendered color frames. It accepts three values:

  • none — No overlay (default).

  • model — Draws wireframes for controllable actor tracks from the model checkpoint. Requires a composite model (scenes with dynamic actors). Raises an error at startup if the loaded model does not support actor tracks.

  • artifact — Draws wireframes for all annotation tracks from sequence_tracks.json in the USDZ artifact. Includes all tracked objects regardless of controllability.

Use --bbox-color R G B to set the wireframe color as three integers (0–255). Default is 0 255 0 (green).

Output types#

By default, the render command produces RGB color frames. Use --output-type to request additional output types. The flag is repeatable, so you can produce multiple output types in a single run:

docker run --shm-size=64g -it --rm --gpus all \
    --network host \
    --volume /path/to/output/folder:/workdir/output \
    nvcr.io/nvidia/nre/nre-ga:26.04 \
    -- render \
    --artifact-path "${{USDZ_PATH}}" \
    --output-dir "${{RENDER_OUTPUT_DIR}}" \
    --output-type color \
    --output-type distance-colorized \
    --output-type semantics-colorized \
    --camera-id camera_front_wide_120fov

The three supported types are:

  • color — Standard RGB render (default).

  • distance-colorized — Depth map colorized by distance from the sensor. The color scale saturates at --colorization-max-distance meters (default: 80.0). Reduce this value to increase contrast for close-range scenes — for example, --colorization-max-distance 20.0 saturates at 20 m.

  • semantics-colorized — Semantic segmentation colorized using the class colors packaged into the USDZ artifact. By default, a semantic_classes.png legend is written alongside the frames. Pass --no-semantics-legend to suppress it.

Each output type writes to its own subdirectory under --output-dir:

Output type

Directory

Additional files

color

color/

—

distance-colorized

distance_colorized/

—

semantics-colorized

semantics_colorized/

semantic_classes.png (legend)

Note: A gRPC interface for distance and semantic outputs is planned for a future release.

More options#

For more details on available options for the render command, pass the --help option.

To render NuRec models through the NuRec gRPC service, see Use the NuRec gRPC API Server.

On this page