Generating HDR Video from SDR Video

2026-05-14 · Source: Takara TLDR - Daily AI Papers · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Emerging Technologies & Innovation · Depth: Expert, medium

Summary

A new framework addresses the persistent challenge of upconverting legacy Standard Dynamic Range (SDR) videos to High Dynamic Range (HDR) format. This framework utilizes large-scale generative video models to synthesize HDR video from casual SDR footage. It introduces a Multi-Exposure Video Model (MEVM) that predicts exposure-bracketed linear SDR video sequences from a single nonlinear SDR input. Additionally, a learnable Video Merging Model (VMM) combines these predicted sequences into a high-quality HDR output, preserving detail in both shadows and highlights. Extensive quantitative and qualitative evaluations, including a user study, confirm the approach's robustness for in-the-wild consumer videos and even iconic films, supporting HDR synthesis pipelines built on existing SDR generative video models.

Key takeaway

For research scientists developing video processing pipelines, this framework offers a robust solution for SDR-to-HDR conversion. You should consider integrating this multi-exposure prediction and merging approach into your generative video model workflows to achieve high-quality HDR outputs, especially for diverse "in-the-wild" footage. This could significantly enhance the visual fidelity of legacy content.

Key insights

A new framework converts SDR video to HDR using generative models and a multi-exposure prediction and merging process.

Principles

• Generative models can synthesize complex video properties.
• Exposure bracketing enhances dynamic range conversion.

Method

The method involves a Multi-Exposure Video Model (MEVM) to predict exposure-bracketed SDR sequences, followed by a Video Merging Model (VMM) to combine them into a high-quality HDR output.

In practice

• Convert casual consumer videos to HDR.
• Upgrade iconic films to HDR format.

Topics

• HDR Video Synthesis
• SDR to HDR Conversion
• Generative Video Models
• Multi-Exposure Video Model
• Video Merging Model

Code references

• MCL-ZJU/FDIM
• google-deepmind/videoprism

Best for: Research Scientist, AI Scientist, Machine Learning Engineer, Computer Vision Engineer

Related on AIssential

• RefDecoder: Enhancing Visual Generation with Conditional Video Decoding — Conditional video decoders improve detail preservation and consistency in visual generation.
• Advancing Digital Twin Generation Through a Novel Simulation Framework and Quantitative Benchmarking — A novel simulation framework enables quantitative benchmarking of 3D photogrammetry pipelines using synthetic ground-truth data.
• Linear Image Generation by Synthesizing Exposure Brackets — Generating linear images as exposure brackets overcomes VAE limitations for high dynamic range content.
• Reshoot-Anything: A Self-Supervised Model for In-the-Wild Video Reshooting — Self-supervised learning from monocular videos enables robust video reshooting by generating pseudo multi-view data.
• RISE-Video: Can Video Generators Decode Implicit World Rules? — RISE-Video evaluates generative video models' reasoning over implicit world rules beyond visual fidelity.
• Building Generative Image & Video models at Scale - Sander Dieleman (Veo and Nano Banana) — Diffusion models generate high-quality audiovisual data through iterative denoising of learned compressed latent representations.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Takara TLDR - Daily AI Papers.