AI Paper Atlas

Research map / Image generation / Fast diffusion sampling

DiffusionLight-Turbo: Accelerated Light Probes for Free via Single-Pass Chrome Ball Inpainting

Worameth Chinchuthakun, Pakkapon Phongthawee, Amit Raj, Varun Jampani, Pramook Khungurn, Supasorn Suwajanakorn

TPAMI 2026 · Image generation › Fast diffusion sampling · approach: Training-free diffusion acceleration

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add Exposure LoRA (ablation)DiffusionLight · Laval Indoorangular error (mirror): 7.053 -> 5.936Improved
add iterative inpainting (ablation)DiffusionLight · Laval Indoorangular error (mirror): 6.146 -> 5.936Improved
add iterative inpainting (ablation)DiffusionLight · Poly Havenangular error (matte): 3.560 -> 3.595No clear effect
add LoRA swapping instead of Turbo-SDEdit (ablation)DiffusionLight-Turbo · Laval Indoortime (mins): 1.229 -> 0.535Improved
add LoRA swapping instead of Turbo-Pred (ablation)DiffusionLight-Turbo · Laval Indoorangular error (mirror): 6.449 -> 6.114Improved
apply Turbo LoRA for speedupDiffusionLight · Laval Indoortime (mins) and angular error (mirror): 31.889 -> 0.535 mins; 5.936 -> 6.114Mixed
use DiffusionLight instead of StyleLightStyleLight (reproduced) · Poly Havenangular error (mirror): 7.122 -> 4.286Improved
use Turbo instead of full DiffusionLightDiffusionLight · Laval Indoor (array of spheres)angular error: 5.25 -> 7.35Hurt
increase balls N from 5 to 30 (k=1) (ablation)DiffusionLight · Poly Havenangular error (mirror): 4.380 -> 4.282Improved
use two iterations rather than one (N=30) (ablation)DiffusionLight · Poly Havenangular error (mirror): 4.282 -> 4.286No clear effect
train Exposure LoRA on timesteps 900-999 rather than 0-999 (ablation)Exposure LoRA · validation mirrorangular error: 6.292 -> 5.464Improved
train Turbo LoRA on timesteps 500-999 rather than 0-999 (ablation)Turbo LoRA · validation mirrorangular error: 5.891 -> 5.539Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy