CVL-UESTC/FR-INR

CVPR 2024-Improved Implicit Neural Representation with Fourier Reparameterized Training

25
/ 100
Experimental

This project helps computer vision researchers and practitioners who use Implicit Neural Representations (INRs) for tasks like image or 3D shape reconstruction. It provides a new training method, Fourier Reparameterization, to improve how INRs learn details, resulting in outputs with richer textures and fewer visual flaws. You input training data (e.g., images, 3D occupancy grids) and get a more accurate and detailed INR model.

No commits in the last 6 months.

Use this if you are working with Implicit Neural Representations and need to overcome issues like blurry outputs or poor detail reproduction, specifically addressing the 'spectral bias' problem.

Not ideal if you are looking for a completely new INR architecture rather than an improved training technique for existing ones.

Implicit Neural Representations Computer Vision 3D Reconstruction Image Regression Neural Rendering
No License Stale 6m No Package No Dependents
Maintenance 2 / 25
Adoption 9 / 25
Maturity 8 / 25
Community 6 / 25

How are scores calculated?

Stars

99

Forks

3

Language

Python

License

Last pushed

May 23, 2025

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/diffusion/CVL-UESTC/FR-INR"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.