[AAAI 2024] Devignet: High-Resolution Vignetting Removal via a Dual Aggregated Fusion Transformer With Adaptive Channel Expansion
-
Updated
Jul 25, 2024 - Python
[AAAI 2024] Devignet: High-Resolution Vignetting Removal via a Dual Aggregated Fusion Transformer With Adaptive Channel Expansion
A simple python program correcting vignetting effects with discrete entropy estimation.
Implementation of state of the art algorithms for uneven illomination correction in whole slide imaging
Lens vignetting correction for DaVinci Resolve built from Adobe’s own lens profiles. Reads the vignetting data embedded in Adobe lens profiles and turns it into a DCTL with per-lens/focal-length presets, one correction slider, and sensor-format selection. No guessing, no manual tuning — the same vignetting model that Lightroom uses, inside Resolve.
To associate your repository with the vignetting-correction-algorithm topic, visit your repo's landing page and select "manage topics."