New HDR reconstruction framework uses diffusion models and gain maps

Researchers have developed DOME-HDR, a new framework for reconstructing High Dynamic Range (HDR) images from multiple exposures. This system first generates a standard dynamic range (SDR) image using a LoRA-adapted latent diffusion model, incorporating structural and color information from under- and over-exposed inputs. The generated SDR image then guides a network called HPGM to predict a gain map, enabling reliable dynamic-range expansion to produce a consistent HDR image. Evaluations on several datasets show DOME-HDR achieving state-of-the-art results in HDR reconstruction quality. AI

IMPACT This research advances image reconstruction techniques, potentially improving visual fidelity in photography and digital imaging applications.

RANK_REASON The cluster contains a research paper detailing a novel technical approach. [lever_c_demoted from research: ic=1 ai=0.7]

Read on arXiv cs.CV →

• DOME-HDR
• HPGM
• LoRA

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →