York-SDCNLab/Ghost-DeblurGAN
This is a lightweight GAN developed for real-time deblurring. The model has a super tiny size and a rapid inference time. The motivation is to boost marker detection in robotic applications, however, you may use it for other applications definitely.
This project helps roboticists and automation engineers improve the reliability of fiducial marker detection in systems like drones and automated guided vehicles. It takes blurry images, often caused by motion, and outputs sharpened versions, making it much easier for computer vision systems to accurately identify markers. This is particularly useful for anyone whose robots rely on visual markers for navigation, localization, or object manipulation in dynamic environments.
No commits in the last 6 months.
Use this if your robotic system frequently encounters motion blur in camera feeds, causing your fiducial marker detection (like AprilTags or ArUcos) to fail or be unreliable.
Not ideal if you are trying to deblur general images (e.g., photos of people or landscapes) without specific fiducial markers, as its training is specialized for robotic marker detection.
Stars
50
Forks
10
Language
Python
License
MIT
Category
Last pushed
Sep 12, 2023
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/computer-vision/York-SDCNLab/Ghost-DeblurGAN"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
cs-chan/Exclusively-Dark-Image-Dataset
Exclusively Dark (ExDARK) dataset which to the best of our knowledge, is the largest collection...
ceciliavision/perceptual-reflection-removal
Single Image Reflection Separation with Perceptual Losses
Srameo/LED
[ICCV 2023] Lighting Every Darkness in Two Pairs: A Calibration-Free Pipeline for RAW Denoising...
plemeri/InSPyReNet
Official PyTorch implementation of Revisiting Image Pyramid Structure for High Resolution...
cidautai/DarkIR
CVPR 2025 DarkIR: Robust Low-Light Image Restoration - State of the art low light deblurring....