LiBrA-Net: Lie-Algebraic Bilateral Affine Fields for Real-Time 4K Video Dehazing

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Wang, Yongcong, Shen, Chengchao, Gao, Guangwei, Wang, Wei, Dai, Pengwen, Lu, Dianjie, Zhang, Guijuan, Zheng, Zhuoran
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918496294666240
author Wang, Yongcong
Shen, Chengchao
Gao, Guangwei
Wang, Wei
Dai, Pengwen
Lu, Dianjie
Zhang, Guijuan
Zheng, Zhuoran
author_facet Wang, Yongcong
Shen, Chengchao
Gao, Guangwei
Wang, Wei
Dai, Pengwen
Lu, Dianjie
Zhang, Guijuan
Zheng, Zhuoran
contents Currently, there is a gap in the field of ultra-high-definition (UHD) video dehazing due to the lack of a benchmark for evaluation. Furthermore, existing video dehazing methods cannot run on consumer-grade GPUs when processing continuous UHD sequences of 3--5 frames at a time. In this paper, we address both issues with a new benchmark and an efficient method. Our key observation is that atmospheric dehazing reduces to a per-pixel affine transform governed by the low-frequency depth field, which can be compactly encoded in bilateral grids whose prediction cost is decoupled from the output resolution. Building on this, we propose LiBrA-Net, which factorizes the spatiotemporal affine field into a spatial--color and a temporal bilateral sub-grid predicted at a fixed low resolution, fuses their coefficients in the $\mathfrak{gl}(3)$ Lie algebra under group-theoretic regularization, maps the result to invertible GL(3) transforms via a Cayley parameterization, and restores high-frequency detail through a lightweight input-guided branch. We further release UHV-4K, the first paired 4K video dehazing benchmark with depth, transmission, and optical-flow annotations on every frame. Across UHV-4K, REVIDE, and HazeWorld, LiBrA-Net sets a new state of the art among compared video dehazing methods while running native 4K at 25 FPS on a single GPU with only 6.12 M parameters. Code and data are available at https://anonymous.4open.science/r/LiBrA-Net-42B8.
format Preprint
id arxiv_https___arxiv_org_abs_2605_11508
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle LiBrA-Net: Lie-Algebraic Bilateral Affine Fields for Real-Time 4K Video Dehazing
Wang, Yongcong
Shen, Chengchao
Gao, Guangwei
Wang, Wei
Dai, Pengwen
Lu, Dianjie
Zhang, Guijuan
Zheng, Zhuoran
Computer Vision and Pattern Recognition
Currently, there is a gap in the field of ultra-high-definition (UHD) video dehazing due to the lack of a benchmark for evaluation. Furthermore, existing video dehazing methods cannot run on consumer-grade GPUs when processing continuous UHD sequences of 3--5 frames at a time. In this paper, we address both issues with a new benchmark and an efficient method. Our key observation is that atmospheric dehazing reduces to a per-pixel affine transform governed by the low-frequency depth field, which can be compactly encoded in bilateral grids whose prediction cost is decoupled from the output resolution. Building on this, we propose LiBrA-Net, which factorizes the spatiotemporal affine field into a spatial--color and a temporal bilateral sub-grid predicted at a fixed low resolution, fuses their coefficients in the $\mathfrak{gl}(3)$ Lie algebra under group-theoretic regularization, maps the result to invertible GL(3) transforms via a Cayley parameterization, and restores high-frequency detail through a lightweight input-guided branch. We further release UHV-4K, the first paired 4K video dehazing benchmark with depth, transmission, and optical-flow annotations on every frame. Across UHV-4K, REVIDE, and HazeWorld, LiBrA-Net sets a new state of the art among compared video dehazing methods while running native 4K at 25 FPS on a single GPU with only 6.12 M parameters. Code and data are available at https://anonymous.4open.science/r/LiBrA-Net-42B8.
title LiBrA-Net: Lie-Algebraic Bilateral Affine Fields for Real-Time 4K Video Dehazing
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2605.11508