nurec-fixer

name	nurec-fixer
description	Use to run NVIDIA DiffusionHarmonizer (public successor to the older Fixer recipes) to enhance, harmonize, evaluate, or fine-tune novel-view frames from NRE / NuRec / 3DGS / NeRF reconstructions. Do NOT use for training the 3D reconstruction itself (use `nre`) or for sensor-to-NCore conversion (use `ncore`).
version	0.5.0
tools	["Shell","Read","Write"]
license	CC-BY-4.0 AND Apache-2.0
compatibility	Linux + NVIDIA GPU Ampere+ (compute capability >= 8.0; A100, A10, L40, H100, RTX 30/40/PRO, B200, GB200), Docker + NVIDIA Container Toolkit, ~120 GB free disk. HF_TOKEN required for the `nvidia/Harmonizer` checkpoints and the base `nvidia/Cosmos-Predict2-0.6B-Text2Image` model; NGC_API_KEY needed when pulling the `nvcr.io/nvidia/pytorch:25.10-py3` base image from nvcr.io. Code at github.com/NVIDIA/harmonizer; runtime is the `harmonizer-cosmos-env` image built from the project `Dockerfile.cosmos`, which layers cosmos-predict2 and the required patches on top of `nvcr.io/nvidia/pytorch:25.10-py3`.
dependencies	["bash","docker","nvidia-container-toolkit","python3","git","huggingface_hub"]
metadata	{"author":"NVIDIA NRS <[e-mail]>","tags":["nurec","fixer","harmonizer","diffusionharmonizer","reconstruction","autonomous-vehicles","post-processing","evaluation"],"upstream":"https://github.com/NVIDIA/harmonizer","hf_model":"https://huggingface.co/nvidia/Harmonizer","hf_dataset":"https://huggingface.co/datasets/nvidia/Harmonizer-Dataset","container":"nvcr.io/nvidia/pytorch:25.10-py3","paper":"https://arxiv.org/abs/2602.24096","product_page":"https://research.nvidia.com/labs/sil/projects/diffusion-harmonizer/","model_license":"NVIDIA Open Model License Agreement"}

NVIDIA DiffusionHarmonizer (NuRec post-processing)

Purpose

Run NVIDIA DiffusionHarmonizer on rendered images from neural reconstructions. DiffusionHarmonizer is a single-step, temporally-aware image diffusion enhancer for NeRF / 3DGS / NuRec-style renderings. It improves realism, reduces reconstruction artifacts, and harmonizes inserted dynamic objects with the surrounding scene.

When to Use / When NOT to Use

Use this skill when the user has rendered frames from NRE, NuRec, 3DGS, NeRF, or a similar reconstruction pipeline and wants to enhance, harmonize, evaluate, or optionally fine-tune the DiffusionHarmonizer model.

Do NOT use this skill when:

• The user wants to train or render the 3D reconstruction itself (use nre).
• The user wants to convert raw sensor data to NCore V4 (use ncore).
• The user wants a generic photo enhancer. DiffusionHarmonizer is tuned for neural-reconstruction artifacts and object-insertion failures.
• The user only wants NRE inline rendering with --enable-difix. That remains an NRE runtime feature; use the nre skill for the complete serve-grpc / render-grpc command shape.

What changed from the older Fixer skill

This skill follows the public NVIDIA/harmonizer release, not the older NGC JIT .pt artifact recipe. Use these public release artifacts:

• Code: https://github.com/NVIDIA/harmonizer
• Model: nvidia/Harmonizer on Hugging Face (the paper checkpoint models/diffusion_harmonizer.pkl), plus the base nvidia/Cosmos-Predict2-0.6B-Text2Image model that inference also requires.
• Checkpoint download: ./download_checkpoints.sh from the repo root. It fetches the Harmonizer checkpoints into models/ (diffusion_harmonizer.pkl, harmonizer_nontemporal.pt) and the base Cosmos DiT + tokenizer into src/checkpoints/nvidia/Cosmos-Predict2-0.6B-Text2Image/.
• Runtime: the harmonizer-cosmos-env image built from Dockerfile.cosmos (base nvcr.io/nvidia/pytorch:25.10-py3).
• Inference entry: src/inference_pix2pix_turbo_harmonizer.py, run from inside /work/src so it can import its sibling modules.