mlsanityData to 14 Oct 2026

Neural Block-wise Auto-exposure for Single-Shot High Dynamic Range Imaging

arXiv 2610.18733 ↗First posted 9 Oct 2026cs.CVcited by 0

Not on the map. Why?

Abstract (arXiv)

Block-wise varying exposures are effective for cap- turing appearance of a real-world high dynamic range (HDR) scene via a single-shot. However, brightness order reversal (BOR) artifacts, blocking artifacts and loss of details in highlight/shadow regions are possible issues for block-wise varying exposures. In this paper, a novel neural block-wise auto-exposure algorithm is proposed to address the issues. The proposed algorithm maps a temporary image and its exposure time to the exposure times of all non-overlapping 16 X 16 blocks. The temporary image can be either an sRGB or a raw image. Although the sRGB image is widely adopted in existing auto-exposure algorithms, exposure times of blocks in the highlight and shadow regions might not be accurately learned due to the limited radiance measurement capability of the sRGB image. This paper thus also explores the feasibility of temporary image as a raw image. Due to the better radiance measurement capability of the raw image, the exposure times of blocks are learned more accurately, and more details in the highlight/shadow regions of the real-world HDR scene are captured by the resultant HDR imaging system. Experiment results validate the proposed neural block-wise auto-exposure algorithm.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 20 of its 47 references that are in the map. First posted on arXiv 9 Oct 2026.

Builds on

20 of 47 references.

References · 20

  1. 2015U-Net: Convolutional Networks for Biomedical Image Segmentationcited by 19,713
  2. 2021MUSIQ: Multi-scale Image Quality Transformercited by 1,081
  3. 2020Single-Image HDR Reconstruction by Learning to Reverse the Camera Pipelinecited by 132
  4. 2019Attention-guided Network for Ghost-free High Dynamic Range Imagingcited by 117
  5. 2021HDR-NeRF: High Dynamic Range Neural Radiance Fieldscited by 114
  6. 2017Deep High Dynamic Range Imaging with Large Foreground Motionscited by 99
  7. 2017DeepFuse: A Deep Unsupervised Approach for Exposure Fusion with Extreme Exposure Image Pairscited by 78
  8. 2022Ghost-free High Dynamic Range Imaging with Context-aware Transformercited by 65
  9. 2020HDR Environment Map Estimation for Real-Time Augmented Realitycited by 47
  10. 2023RawHDR: High Dynamic Range Image Reconstruction from a Single Raw Imagecited by 32
  11. 2025UltraFusion: Ultra High Dynamic Imaging using Exposure Fusioncited by 21
  12. 2020End-to-End Differentiable Learning to HDR Image Synthesis for Multi-exposure Imagescited by 15
8 more
  1. 2020Single Image Brightening via Multi-Scale Exposure Fusion with Hybrid Learningcited by 12
  2. 2021Unpaired Learning for High Dynamic Range Image Tone Mappingcited by 11
  3. 2021Deep Joint Demosaicing and High Dynamic Range Imaging within a Single Shotcited by 10
  4. 2023Self-Supervised High Dynamic Range Imaging with Multi-Exposure Images in Dynamic Scenescited by 9
  5. 2024Learning to Control Camera Exposure via Reinforcement Learningcited by 9
  6. 2025AdaptiveAE: An Adaptive Exposure Strategy for HDR Capturing in Dynamic Scenescited by 3
  7. 2023Examining Autoexposure for Challenging Scenescited by 2
  8. 2026Multi-exposure HDR Imaging: A Review of Pixel-level and Feature-level Reconstruction Methodscited by 1

Built on by

Not cited yet.