--- title: Scale AV Perception Across Vehicle Platforms with NVIDIA... description: NVIDIA Omniverse NuRec enables autonomous vehicle developers to adapt perception stacks to new vehicle carlines without collecting fresh real-world datasets.... canonical: https://daily.dev/posts/scale-av-perception-across-vehicle-platforms-with-nvidia-omniverse-nurec-vfuubxqw6 twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Scale AV Perception Across Vehicle Platforms with NVIDIA Omniverse NuRec | daily.dev og:description: NVIDIA Omniverse NuRec enables autonomous vehicle developers to adapt perception stacks to new vehicle carlines without collecting fresh real-world datasets.... og:url: https://daily.dev/posts/scale-av-perception-across-vehicle-platforms-with-nvidia-omniverse-nurec-vfuubxqw6 og:image: https://api.daily.dev/og/posts/vFuUBxqW6.png og:image:alt: Scale AV Perception Across Vehicle Platforms with NVIDIA Omniverse NuRec og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # Scale AV Perception Across Vehicle Platforms with NVIDIA Omniverse NuRec **[NVIDIA Developer](https://daily.dev/sources/nvidiadev)** · 14 min read · 0 upvotes · 0 comments ## Summary NVIDIA Omniverse NuRec enables autonomous vehicle developers to adapt perception stacks to new vehicle carlines without collecting fresh real-world datasets. Using 3D Gaussian splatting, NuRec reconstructs existing recorded drives and renders novel camera views matching a target vehicle's sensor rig configuration. The tutorial walks through downloading a reconstructed scene from the Physical AI NuRec Dataset, rendering target-rig views, refining frames with NVIDIA Harmonizer to fix neural-rendering artifacts, and training a perception model on the resulting synthetic data. It also covers converting raw drive recordings into the NCore format and generating auxiliary data (segmentation, depth, ego masks) needed for reconstruction. NVIDIA reports using this workflow internally for a program adapting to a new camera configuration before target-carline data was available. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### How can I adapt an autonomous vehicle perception model to a new sensor rig without collecting new real-world data for that carline? NVIDIA Omniverse NuRec reconstructs existing recorded drives using 3D Gaussian splatting and renders new camera views matching a target vehicle's sensor rig. The workflow involves downloading a reconstructed scene, exporting a custom rig trajectory, rendering target camera views, refining frames with NVIDIA Harmonizer to remove neural-rendering artifacts, then training the perception model on this synthetic output. _daily.dev surfaces workflows like NuRec for teams adapting perception stacks across vehicle platforms._ ### What does NVIDIA Harmonizer do to NuRec-rendered camera frames? NVIDIA Harmonizer is a public, temporally aware post-processing model that corrects view-dependent artifacts, inconsistent color or tone, and poorly reconstructed dynamic objects in NuRec and similar neural renderings. It improves visual quality but does not fix incorrect calibration, recover unreconstructed scene coverage, or replace target-camera validation. _engineers refining synthetic AV training data can track tools like Harmonizer through daily.dev._ ### What metadata files are required to reconstruct a driving scene with NVIDIA NuRec's NCore format? Three Parquet files are required alongside synchronized camera videos: calibration_estimate.parquet (camera lens parameters and mounting positions), egomotion_estimate.parquet (vehicle position and orientation over time), and object_fused.parquet (moving objects, classes, dimensions, and track IDs). These map into NCore V4 components like IntrinsicsComponent, PosesComponent, and CuboidsComponent via a companion converter. _daily.dev helps AV engineers keep up with data pipeline requirements for tools like NCore and NuRec._ ## Similar posts on daily.dev - [Optimizing a Neural Reconstruction Pipeline Using NVIDIA Nsight Developer Tools](https://daily.dev/posts/optimizing-a-neural-reconstruction-pipeline-using-nvidia-nsight-developer-tools-kgoma4nfs) · NVIDIA Developer · 0 upvotes · 0 comments - [Reconstruct a Scene in NVIDIA Isaac Sim Using Only a Smartphone](https://daily.dev/posts/reconstruct-a-scene-in-nvidia-isaac-sim-using-only-a-smartphone-omert0rjc) · NVIDIA Developer · 0 upvotes · 0 comments - [Build Synthetic Data Pipelines to Train Smarter Robots with NVIDIA Isaac Sim](https://daily.dev/posts/build-synthetic-data-pipelines-to-train-smarter-robots-with-nvidia-isaac-sim-1a03g4zbf) · NVIDIA Developer · 0 upvotes · 0 comments - [Into the Omniverse: NVIDIA GTC Showcases Virtual Worlds Powering the Physical AI Era](https://daily.dev/posts/into-the-omniverse-nvidia-gtc-showcases-virtual-worlds-powering-the-physical-ai-era-z0vlseat7) · NVIDIA · 0 upvotes · 0 comments --- [View this post on daily.dev](https://daily.dev/posts/scale-av-perception-across-vehicle-platforms-with-nvidia-omniverse-nurec-vfuubxqw6) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"Scale AV Perception Across Vehicle Platforms with NVIDIA Omniverse NuRec","url":"https://daily.dev/posts/scale-av-perception-across-vehicle-platforms-with-nvidia-omniverse-nurec-vfuubxqw6","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/scale-av-perception-across-vehicle-platforms-with-nvidia-omniverse-nurec-vfuubxqw6"},"datePublished":"2026-08-31T16:02:16.098Z","dateModified":"2026-08-31T16:02:40.669Z","description":"NVIDIA Omniverse NuRec enables autonomous vehicle developers to adapt perception stacks to new vehicle carlines without collecting fresh real-world datasets....","image":"https://media.daily.dev/image/upload/s--CxzD6vbw--/f_auto/v1722860399/public/Placeholder%2005","thumbnailUrl":"https://media.daily.dev/image/upload/s--CxzD6vbw--/f_auto/v1722860399/public/Placeholder%2005","isAccessibleForFree":true,"articleSection":"NVIDIA Developer","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"NVIDIA Developer","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/86e45aab42ba48ce83103d01b1119910","url":"https://daily.dev/sources/nvidiadev"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/scale-av-perception-across-vehicle-platforms-with-nvidia-omniverse-nurec-vfuubxqw6","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"","timeRequired":"PT14M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"NVIDIA Developer","item":"https://daily.dev/sources/nvidiadev"},{"@type":"ListItem","position":3,"name":"Scale AV Perception Across Vehicle Platforms with NVIDIA Omniverse NuRec"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/scale-av-perception-across-vehicle-platforms-with-nvidia-omniverse-nurec-vfuubxqw6#faq","mainEntity":[{"@type":"Question","name":"How can I adapt an autonomous vehicle perception model to a new sensor rig without collecting new real-world data for that carline?","acceptedAnswer":{"@type":"Answer","text":"NVIDIA Omniverse NuRec reconstructs existing recorded drives using 3D Gaussian splatting and renders new camera views matching a target vehicle's sensor rig. The workflow involves downloading a reconstructed scene, exporting a custom rig trajectory, rendering target camera views, refining frames with NVIDIA Harmonizer to remove neural-rendering artifacts, then training the perception model on this synthetic output. daily.dev surfaces workflows like NuRec for teams adapting perception stacks across vehicle platforms."}},{"@type":"Question","name":"What does NVIDIA Harmonizer do to NuRec-rendered camera frames?","acceptedAnswer":{"@type":"Answer","text":"NVIDIA Harmonizer is a public, temporally aware post-processing model that corrects view-dependent artifacts, inconsistent color or tone, and poorly reconstructed dynamic objects in NuRec and similar neural renderings. It improves visual quality but does not fix incorrect calibration, recover unreconstructed scene coverage, or replace target-camera validation. engineers refining synthetic AV training data can track tools like Harmonizer through daily.dev."}},{"@type":"Question","name":"What metadata files are required to reconstruct a driving scene with NVIDIA NuRec's NCore format?","acceptedAnswer":{"@type":"Answer","text":"Three Parquet files are required alongside synchronized camera videos: calibration_estimate.parquet (camera lens parameters and mounting positions), egomotion_estimate.parquet (vehicle position and orientation over time), and object_fused.parquet (moving objects, classes, dimensions, and track IDs). These map into NCore V4 components like IntrinsicsComponent, PosesComponent, and CuboidsComponent via a companion converter. daily.dev helps AV engineers keep up with data pipeline requirements for tools like NCore and NuRec."}}]} ```