AI Paper Atlas

Research map / Image generation / Editing and style transfer

Deep CG2Real: Synthetic-to-Real Translation via Image Disentanglement

Sai Bi, Kalyan Sunkavalli, Federico Perazzi, Eli Shechtman, Vladimir G. Kim, Ravi Ramamoorthi

ICCV 2019 · Image generation › Editing and style transfer · approach: Image-to-image translation · 35 citations (OpenAlex)

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
predict shading and multiply by albedo, with shading discriminator, instead of directly predicting PBR image from OpenGL (ablation)OpenGL-to-PBR stage · FID O->PFID: 21.01 -> 7.40Improved
predict shading (with shading discriminator) rather than PBR image directly from albedo, normal, shading input (ablation)OpenGL-to-PBR stage · FID O->PFID: 11.63 -> 7.40Improved
add albedo, normal and shading buffers as generator input (ablation)p2pHD · FID O->PFID: 21.01 -> 11.63Improved
translate albedo and shading with separate generators instead of image space CycleGAN (ablation)PBR-to-real stage · FID P->RFID: 54.80 -> 53.48Improved
use two-stage pipeline instead of single-stage CycleGANOpenGL-to-real translation · FID O->RFID: 59.42 -> 53.48Improved
use two-stage pipeline instead of single-stage T2NetOpenGL-to-real translation · FID O->RFID: 65.33 -> 53.48Improved
train normal estimator on PBR-translated images instead of OpenGL images (ablation)normal estimation network · NYUv2 normal estimationmean angle error (deg): 37.73 -> 34.82Improved
add PBR-to-real translation after first stage for normal estimation (ablation)normal estimation network · NYUv2 normal estimationmean angle error (deg): 34.82 -> 33.15Improved
use disentangled PBR-to-real instead of image-space CycleGAN for normal estimation (ablation)PBR-to-real stage · NYUv2 normal estimationmean angle error (deg): 33.90 -> 33.15Improved
use two-stage translation instead of direct OpenGL-to-real CycleGAN for normal estimationOpenGL-to-real translation · NYUv2 normal estimationmean angle error (deg): 36.33 -> 33.15Improved
train depth estimator on PBR-translated images instead of OpenGL images (ablation)depth estimation network · NYUv2 depth estimationRMSE: 1.0770 -> 1.0293Improved
add PBR-to-real translation for depth estimation (ablation)depth estimation network · NYUv2 depth estimationRMSE: 1.0293 -> 0.9774Improved

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