circle-group/hktex
Heat Kernel Textures: the Geodesic Gaussians That Do Not Splat
Older snapshot: check the original source for current information.
From the publisher
README & documentation
Source previewRead the project’s overview, installation instructions and usage examples. The original README is the source of truth.
Simone Foti* · Caner Korkmaz* · Stefanos Zafeiriou · Tolga Birdal * Equal contribution ECCV 2026 (Best Paper and Long Oral) [](https://arxiv.org/abs/2609.07557) [](https://doi.org/10.1007/978-3-032-37595-717) [](https://circle-group.github.io/research/HeatKernelTextures/) Heat Kernel Textures (HKTex) are an intrinsic texture representation for triangular meshes. Instead of relying on a UV atlas, HKTex represents appearance with anisotropic heat kernels—the geodesic counterparts of Gaussians—defined directly on the surface. Kernel positions, optimisation, pruning, and densification all operate on the mesh. HKTex avoids UV seams, distortion, wasted atlas space, duplicated vertices, and uneven texel resolution. The representation can be fitted from an existing texture or from multi-view images and integrates with physically based rendering. The environments below target Python 3.11, Linux, and an NVIDIA GPU with a CUDA 12.9-compatible driver. Mamba is recommended for environment management. To create the environment, open a terminal and run the the following commands to install the necessary dependencies: This optional environment includes the tiny-cuda-nn dependency used…
Read the full README ↗ · Preview checked 2026-10-30T10:41:28.753Z
Inside the original README — Document outline
- Heat Kernel Textures: the Geodesic Gaussians That Do Not Splat
- Highlights
- Installation
- HKTex
- Neural texture baselines
- Quick start
- Configurations and experiments
- Repository structure
- Citation
- Links
- License
Headings are captured from the source. Links open the publisher’s document, not a locally hosted copy.
Links from the README
References supplied by the publisher, not independently verified endorsements. Check the destination before downloading files or entering credentials.
Documentation belongs to its respective authors. Reported project/model license: MIT. A listing is not a grant of reuse or training rights. Confirm the document’s own terms at the source.
What this repository does
Heat Kernel Textures: the Geodesic Gaussians That Do Not Splat
Repository facts
- Owner
- circle-group
- Primary language
- Python
- Stars
- 140
- Forks
- 3
- Open issues + pull requests
- 0
- License
- MIT
- Archived
- No
- Default branch
- main
- Created
- 2026-09-09T14:48:46.000Z
- Last push
- 2026-09-10T14:29:28.000Z
Topics and intended use
No topics were included in the latest source metadata.
Review the README for scope, installation, examples and limitations. We do not run repository code or certify it.
Evaluate before installing
Review licensing and dependencies, inspect recent commits and unresolved issues, and test in an isolated environment before production use. Stars and forks alone cannot answer these questions.
Source and freshness
Source: GitHub. Metadata observed 2026-09-16T06:39:39.411Z. Daily imports are snapshots, not real-time monitoring.
Popularity and source listings do not establish security, suitability, licensing rights or benchmark performance.
Related repositories
EbookFoundation/free-programming-books
donnemartin/system-design-primer