AI Paper Atlas

Research map / Image generation / GANs

COCO-GAN: Generation by Parts via Conditional Coordinating

Chieh Hubert Lin, Chia-Che Chang, Yu-Sheng Chen, Da-Cheng Juan, Wei Wei, Hwann-Tzong Chen

ICCV 2019 · Image generation › GANs · approach: Conditional GANs · 134 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
use discrete rather than continuous coordinate sampling (ablation)COCO-GAN (N2,M2,S16) · CelebA 64x64FID: 6.13 -> 4.87Improved
let the discriminator see the full image (optimal D) (ablation)COCO-GAN · CelebA 64x64FID: 4.87 -> 4.05Improved
let the generator produce the full image (optimal G) (ablation)COCO-GAN · CelebA 64x64FID: 4.87 -> 6.12No clear effect
use one generator per spatial coordinate instead of a shared conditional generator (ablation)COCO-GAN · CelebA 64x64FID: 4.87 -> 7.26Hurt
generate and discriminate by parts rather than full images (COCO-GAN vs projection D backbone)Projection D · CelebA 128x128FID: 19.55 -> 5.74Improved
use COCO-GAN generation by parts reducing macro patch to part of imageProjection D backbone · CelebA 128x128 (batch size 128)training memory (MB): 17,184 MB -> 8,992 MB (47.7% reduction)Improved
shrink micro patch size down to 4x4 pixels (ablation)COCO-GAN · CelebA 128x128FID: overall FID slightly degradesHurt
use a macro patch smaller than 1/16 of the full image (32x32 macro, 16x16 micro) (ablation)COCO-GAN · CelebA 128x128image qualityHurt
add CoordConv to COCO-GAN (ablation)COCO-GAN · CelebAvisual qualityNo clear effect
freeze later generator layers and train only the first two layers during post-training (ablation)COCO-GAN post-training · LSUN 256x256 beyond-boundaryartifact stabilityImproved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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