A high dynamic range illumination estimation method, device and storage medium

By inserting virtual light probes into the video and utilizing an HDR-aware variational autoencoder and a video diffusion model, the problems of temporal consistency and HDR lighting output in video lighting estimation are solved, achieving temporally consistent high dynamic range lighting estimation, which is suitable for complex scenes and fast motion.

CN121685353BActive Publication Date: 2026-05-19OPENBAYES (TIANJIN) IT CO LTD +1
0 Cites 0 Cited by

Patent Information

Application Number
CN202610186659.3
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2026-02-10
Publication Date
2026-05-19
Estimated Expiration
2046-02-10

AI Technical Summary

Technical Problem

Existing illumination estimation methods lack temporal consistency between frames when processing video, resulting in flickering. Furthermore, existing video generation models cannot directly output physically realistic high dynamic range illumination information.

Method used

The illumination estimation task is refactored into an in-video rendering task. By inserting virtual light probes into the video, physically reliable reflection information is generated using the spatiotemporal context of the video generation model. Illumination estimation is performed using an HDR-aware variational autoencoder and a video diffusion model. Combined with a learnable tone mapping adapter and a low-rank adaptation strategy, temporally consistent HDR environment maps are generated.

Benefits of technology

It achieves temporally consistent high dynamic range illumination estimation in complex scenes, avoids flickering issues, and directly outputs physically realistic HDR illumination information, adapting to fast camera movement and complex occlusion scenes.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The present application relates to a kind of high dynamic range illumination estimation method, device and storage medium, utilize the spatiotemporal generation prior of video generation model, by constructing "in-video drawing" task, from ordinary low dynamic range video sequence, restore time sequence consistent and physically realistic high dynamic range environment map.The input video is applied mask and combined with specific text prompt, guide model to generate virtual light probe containing surrounding environment reflection information in picture center;Variational autoencoder of high dynamic range perception is constructed, by introducing learnable tone mapping and inverse tone mapping adapter, so that video diffusion model can directly understand and output high dynamic range illumination characteristics in logarithmic domain latent space.The present application can output high-quality dynamic high dynamic range environment sequence, when processing dynamic scene, fast camera motion and complex occlusion, it has stronger robustness, applicable to augmented reality, virtual object implantation and scene relighting etc.Application.
Need to check novelty before this filing date? Find Prior Art