weihaox/awesome-neural-rendering

Resources of Neural Rendering

What It Is

An "awesome list" — a hand-picked collection of links to papers, code, and projects on the topic of neural rendering. Neural rendering combines computer graphics and machine learning to create or edit images and video from 3D models and neural networks.

The project has no code or software of its own; it is purely a reference hub, organized into topic-linked markdown files in the docs/ folder.

What Topics It Covers

From the table of contents, the list is divided into 12 categories, each linking to a dedicated page:

  • Introduction and surveys (getting started, overviews)
  • Inverse rendering (de-rendering) — extracting scene properties from images
  • Neural rerendering — re-generating images using neural methods
  • Differentiable rendering — making rendering pipelines trainable
  • Implicit neural representation and rendering — representing scenes as neural functions
  • Volumetric performance capture (free viewpoint video) — capturing 3D video from cameras
  • Semantic photo synthesis and manipulation — editing photos based on semantic labels
  • Texture and surface embedding or mapping — attaching neural textures to surfaces
  • Novel-view synthesis — generating new camera angles from existing images
  • Light, reflectance, illuminance, and shade — understanding and recreating lighting
  • Dubbing and talking-head animation — generating mouth movements and head motion for speech
  • Motion transfer, retargeting, and reenactment — transferring movement between people or characters

How It's Organized

Each category is a separate markdown file (e.g., docs/DIFFERENTIABLE-RENDERING.md). The main README links to those files. Entry items inside the files are formatted in a standard pattern:

  • Paper name (bold)
  • Author names with links
  • Conference/journal and year, plus optional links to PDF, project page, code, video, and dataset

How to Contribute

The README invites anyone to contribute: if you spot a missing paper or have suggestions, you can open a pull request on GitHub. Contributors can either manually edit entries or use a script (linked in the README) to generate the markdown in the prescribed format.

The project carries the "Awesome" badge (a common mark for high-quality curated lists). Note: several badges (maintenance status, PRs-welcome, license) appear only as comments in the README and are not active.

Quick Summary

  • Type: Curated research resource list (no code)
  • Scope: Neural rendering — all sub-topics listed above
  • Format: Markdown files, one per topic, linked from the main README
  • Community: Open to pull-request contributions with a standard entry template
  • Status: The README advertises it as an awesome list, but does not state how frequently it is updated

Related

  • Project
    isl-org/Open3D-ML

    An extension of Open3D that provides tools, pretrained models, and pipelines for 3D machine learning tasks like semantic segmentation and object detection.

  • Project
    wesbos/3d-model-anything

    A toolkit for capturing a room on a LiDAR-equipped iPhone, then using an AI coding agent guided by a 'scan-to-threejs' skill to reconstruct it as an editable, browsable 3D scene — with a demo office included and an iterative, comparison-checked workflow rather than a one-click solution.

  • Project
    hawk86104/three-vue-tres

    An AI-collaborative Web 3D engineering ecosystem built on Three.js and Vue 3 for creating production-ready digital twins and industrial visualizations.

  • Project
    nv-tlabs/3dgrut

    A hybrid 3D rendering framework that combines Gaussian rasterization and ray tracing to support distorted cameras and complex light effects like reflections and shadows.