Logo

BUPTMMLab

Home Publications Tags Members About
BUPTMMLab
ICCV

From Abyssal Darkness to Blinding Glare: A Benchmark on Extreme Exposure Correction in Real World

Bo Wang, Huiyuan Fu, Zhiye Huang, Siru Zhang, Xin Wang, Huadong Ma

Exposure Correction

View All
From Abyssal Darkness to Blinding Glare: A Benchmark on Extreme Exposure Correction in Real World
ICCV

From Abyssal Darkness to Blinding Glare: A Benchmark on Extreme Exposure Correction in Real World

This paper introduces Real-world Extreme Exposure Dataset (REED) to improve extreme exposure correction in real world scenarios. The method is based on burst capturing with a range of exposures and accurate SIFT-based image alignment. The paper also introduces a method (CLIER) for extreme exposure correction based on luminance normalization, semantic awareness, diffusion, and iterative refinement. The experiments validate the efficacy of the proposed method.

Learning Exposure Correction in Dynamic Scenes
ACM Multimedia Oral

Learning Exposure Correction in Dynamic Scenes

This paper constructs the first real-world paired video dataset DIME, including both underexposure and overexposure dynamic scenes, and proposes an end-to-end video exposure correction network, in which a dual-stream module is designed to deal with both underexposure and overexposure factors.

Region-Aware Exposure Consistency Network for Mixed Exposure Correction
AAAI

Region-Aware Exposure Consistency Network for Mixed Exposure Correction

An effective Region-aware Exposure Correction Network (RECNet) is introduced that can handle mixed exposure by adaptively learning and bridging different regional exposure representations and an exposure contrastive regularization strategy under the constraints of intra-regional exposure consistency and inter-regional exposure continuity is proposed.

Lowlight Enhancement

View All
Exploring in Extremely Dark: Low-Light Video Enhancement with Real Events
ACMMM

Exploring in Extremely Dark: Low-Light Video Enhancement with Real Events

This paper proposes the Real-Event Embedded Network (REN) for low-light video enhancement using real events to restore details in extremely dark areas. It introduces the Event-Image Fusion module and unsupervised losses for semi-supervised training on unpaired data. Experiments demonstrate superiority over state-of-the-art methods.

Dancing in the Dark: A Benchmark towards General Low-light Video Enhancement
ICCV

Dancing in the Dark: A Benchmark towards General Low-light Video Enhancement

This paper introduces a high-quality low-light video dataset (DID) and a Retinex-based method called Light Adjustable Network (LAN) for general low-light video enhancement. The dataset features dynamic videos with multiple exposures and cameras, while LAN iteratively refines illumination for adaptive enhancement.

You Do Not Need Additional Priors or Regularizers in Retinex-based Low-light Image Enhancement
CVPR

You Do Not Need Additional Priors or Regularizers in Retinex-based Low-light Image Enhancement

This paper proposes a regularizer-free Retinex decomposition and synthesis network (RFR) for low-light image enhancement. It introduces a contrastive learning method and a self-knowledge distillation method to train the model without additional priors or regularizers. The approach extracts reflectance and illumination features and synthesizes them end-to-end, achieving superior performance on various datasets.

Recent Publications

From Abyssal Darkness to Blinding Glare: A Benchmark on Extreme Exposure Correction in Real World

From Abyssal Darkness to Blinding Glare: A Benchmark on Extreme Exposure Correction in Real World

Jun 26, 2025

This paper introduces Real-world Extreme Exposure Dataset (REED) to improve extreme exposure...
exposure correction iccv feature
Exploring in Extremely Dark: Low-Light Video Enhancement with Real Events

Exploring in Extremely Dark: Low-Light Video Enhancement with Real Events

Oct 28, 2024

This paper proposes the Real-Event Embedded Network (REN) for low-light video enhancement using real...
lowlight enhancement acm multimedia event
Learning Exposure Correction in Dynamic Scenes

Learning Exposure Correction in Dynamic Scenes

Oct 28, 2024

This paper constructs the first real-world paired video dataset DIME, including both underexposure...
exposure correction acm multimedia
Region-Aware Exposure Consistency Network for Mixed Exposure Correction

Region-Aware Exposure Consistency Network for Mixed Exposure Correction

Feb 28, 2024

An effective Region-aware Exposure Correction Network (RECNet) is introduced that can handle mixed...
exposure correction aaai
Dancing in the Dark: A Benchmark towards General Low-light Video Enhancement

Dancing in the Dark: A Benchmark towards General Low-light Video Enhancement

Oct 4, 2023

This paper introduces a high-quality low-light video dataset (DID) and a Retinex-based method called...
lowlight enhancement iccv dataset
You Do Not Need Additional Priors or Regularizers in Retinex-based Low-light Image Enhancement

You Do Not Need Additional Priors or Regularizers in Retinex-based Low-light Image Enhancement

Jun 20, 2023

This paper proposes a regularizer-free Retinex decomposition and synthesis network (RFR) for...
lowlight enhancement cvpr retinex
Page 1 / 1