The marketing name is "portrait lighting" or "relighting"; the filing name is photo relighting using deep neural networks. Google's granted patent US11776095B2, issued October 3, 2023 and naming inventors including Jonathan Barron, claims changing the lighting in a photo after it was taken, using neural networks plus a confidence-learning step. Its CPC tags pair image-enhancement (G06T 5/008) with rendering-illumination (G06T 15/506).
On the record, this is the moment computational photography crosses into the generative. Earlier pipeline steps — fusion, tone mapping, deghosting — reconstruct a faithful image from real captured frames. Relighting goes further: it synthesizes lighting that was never physically present, inferring the 3D shape and reflectance of a face to relight it plausibly. That is generation, not correction.
Why the confidence-learning detail matters: a relighting model that guesses wrong produces uncanny, obviously-fake results. The confidence component lets the system know where its estimate is reliable and where it should hold back — the difference between a usable feature and an artifact generator. The claim's specificity is in that reliability machinery.
Novel, or just renamed? Relighting research predates this grant, but moving it into a shipping computational-photography pipeline with confidence estimation is the claimed advance. The category — neural image manipulation — is hot, and the line between "enhancement" and "generation" is exactly where the interesting IP now sits.
Scope, stated carefully: this is a granted patent to Google on a specific neural-relighting method, within a fast-moving landscape of generative image techniques from many players. It marks where Google's pipeline went generative, not a fence around relighting.
Follow the filing, not the gallery. When a phone offers to change the lighting on a portrait after the fact, it is running something like this 2023 grant — a neural network synthesizing light that was never there. Computational photography has become, in part, computational invention.