Papers › Instance-Conditioned GAN

Instance-Conditioned GAN

10 Sep 2021NeurIPS 2021 12arXiv:2109.05070archive 2025-07-28

Arantxa Casanova, Marlène Careil, Jakob Verbeek, Michal Drozdzal, Adriana Romero-Soriano

Generative Adversarial Networks (GANs) can generate near photo realistic images in narrow domains such as human faces. Yet, modeling complex distributions of datasets such as ImageNet and COCO-Stuff remains challenging in unconditional settings. In this paper, we take inspiration from kernel density estimation techniques and introduce a non-parametric approach to modeling distributions of complex datasets. We partition the data manifold into a mixture of overlapping neighborhoods described by a datapoint and its nearest neighbors, and introduce a model, called instance-conditioned GAN (IC-GAN), which learns the distribution around each datapoint. Experimental results on ImageNet and COCO-Stuff show that IC-GAN significantly improves over unconditional models and unsupervised data partitioning baselines. Moreover, we show that IC-GAN can effortlessly transfer to datasets not seen during training by simply changing the conditioning instances, and still generate realistic images. Finally, we extend IC-GAN to the class-conditional case and show semantically controllable generation and competitive quantitative results on ImageNet; while improving over BigGAN on ImageNet-LT. Code and trained models to reproduce the reported results are available at https://github.com/facebookresearch/ic_gan.

PaperPDFConference PDFCode

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

facebookresearch/ic_gan officialmentioned in papermentioned on GitHubpytorchNOASSERTION report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Conditional Image GenerationImage GenerationUnconditional Image Generation

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Conditional Image Generation ImageNet 128x128 IC-GAN + DA FID 9.5 #15 of 22 Archive leaderboard report
Conditional Image Generation ImageNet 128x128 IC-GAN + DA Inception score 108.6 #15 of 22 Archive leaderboard report
Conditional Image Generation ImageNet 256x256 BigGAN+ [Brock et al.] (chx96) FID 8.1 #4 of 5 Archive leaderboard report
Conditional Image Generation ImageNet 256x256 BigGAN+ [Brock et al.] (chx96) Inception score 144.2 #4 of 5 Archive leaderboard report
Conditional Image Generation ImageNet 256x256 IC-GAN (chx96) + DA FID 8.2±0.1 #5 of 5 Archive leaderboard report
Conditional Image Generation ImageNet 256x256 IC-GAN (chx96) + DA Inception score 173.8±0.9 #5 of 5 Archive leaderboard report
Conditional Image Generation ImageNet 64x64 IC-GAN + DA FID 6.7 #1 of 4 Archive leaderboard report
Conditional Image Generation ImageNet 64x64 IC-GAN + DA Inception score 45.9±0.3 #1 of 4 Archive leaderboard report
Conditional Image Generation ImageNet 64x64 BigGAN* [Brock et al.] +DA FID 10.2±0.1 #4 of 4 Archive leaderboard report
Conditional Image Generation ImageNet 64x64 BigGAN* [Brock et al.] +DA Inception score 30.1±0.1 #4 of 4 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

1x1 ConvolutionAdamBatch NormalizationConditional Batch NormalizationConvolutionDense ConnectionsEarly StoppingFeedforward NetworkGAN Hinge LossLinear LayerNon-Local BlockNon-Local OperationOff-Diagonal Orthogonal RegularizationProjection DiscriminatorReLUResidual BlockResidual ConnectionSAGANSoftmaxSpectral NormalizationTTURTruncation Trick

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections