Story

On air. The free wire runs 15 minutes behind live. See pricing

World, crypto, finance and tech news, scored for impact.

TechHacker NewsFiled 5 days ago

Physically Controllable Relighting of Photographs

Open the sourceReported by Hacker News, yaksoy.github.io, 11 November 2026

Not investment advice. Muhabbit scores how much a story may matter, not which way to trade.

Simon Fraser University researchers Chris Careaga and Yağız Aksoy have unveiled a physically controllable photo relighting method, presented at SIGGRAPH 2025 with a patent pending. The technique infers a colored 3D mesh from a single photograph using monocular geometry and intrinsic-component estimates, letting users place spot lights, point lights, or environmental illumination in the scene before a path-tracing engine and a feed-forward neural renderer produce a photorealistic result. A differentiable rendering process reconstructs in-the-wild illumination, enabling self-supervised training on raw photograph collections without ground-truth relit pairs. The work matters because it brings the explicit, physically based lighting control long standard in 3D tools such as Blender to ordinary photographs, a capability the team frames as a step beyond prior intrinsic-decomposition methods that assume single-color, Lambertian illumination. Licensing inquiries are directed to SFU's Technology Licensing Office and Aksoy. Next to watch: SIGGRAPH presentation details and whether commercial imaging vendors license the method for editing pipelines.

Muhabbit's AI wrote this summary in its own words from Hacker News's report. It can get details wrong, so check the source.

Named in this story

SIGGRAPH, SFU, Blender, PetaPixel, Two Minute Papers