Learning Continuous Exposure Value Representations for Single-Image HDR Reconstruction

TL;DR

This paper introduces CEVR, a novel method for HDR image reconstruction from a single LDR image that uses continuous exposure value representations to generate diverse LDR stacks, improving HDR quality.

Contribution

The paper proposes the continuous exposure value representation (CEVR) using an implicit function, enabling generation of LDR images with arbitrary EVs, including unseen ones, for better HDR reconstruction.

Findings

CEVR improves HDR reconstruction quality over existing methods.

The model can generate LDR images with arbitrary EVs, including unseen EVs.

Experimental results demonstrate superior performance of CEVR.

Abstract

Deep learning is commonly used to reconstruct HDR images from LDR images. LDR stack-based methods are used for single-image HDR reconstruction, generating an HDR image from a deep learning-generated LDR stack. However, current methods generate the stack with predetermined exposure values (EVs), which may limit the quality of HDR reconstruction. To address this, we propose the continuous exposure value representation (CEVR), which uses an implicit function to generate LDR images with arbitrary EVs, including those unseen during training. Our approach generates a continuous stack with more images containing diverse EVs, significantly improving HDR reconstruction. We use a cycle training strategy to supervise the model in generating continuous EV LDR images without corresponding ground truths. Our CEVR model outperforms existing methods, as demonstrated by experimental results.