Research map / Image generation / GANs
FakeCLR: Exploring Contrastive Learning for Solving Latent Discontinuity in Data-Efficient GANs
Ziqiang Li, Chaoyue Wang, Heliang Zheng, Jing Zhang, Bin Li
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 changed | Where | Measured | Effect |
|---|---|---|---|
| add real-image instance discrimination (ablation) | StyleGAN2-ADA-Linear · FFHQ-100 | FID: 82 -> 107 | Hurt |
| add real-image instance discrimination (ablation) | StyleGAN2-ADA-Linear · FFHQ-5K | FID: 9.39 -> 8.50 | Improved |
| add fake-image instance discrimination (ablation) | StyleGAN2-ADA-Linear · FFHQ-100 | FID: 82 -> 76.6 | Improved |
| add fake-image instance discrimination (ablation) | StyleGAN2-ADA-Linear · FFHQ-2K | FID: 13.01 -> 12.69 | Improved |
| add instance-perturbation contrastive loss (ablation) | StyleGAN2-ADA-Linear · FFHQ-100 | FID: 82 -> 43.35 | Improved |
| add instance-perturbation contrastive loss (ablation) | StyleGAN2-ADA-Linear · FFHQ-2K | FID: 13.01 -> 11.29 | Improved |
| add noise-related latent augmentation (ablation) | FakeCLR (instance-perturbation) · FFHQ-100 | FID: 43.35 -> 44.10 | Hurt |
| add noise-related latent augmentation (ablation) | FakeCLR (instance-perturbation) · FFHQ-2K | FID: 11.29 -> 10.65 | Improved |
| add forgetting factor of queue (ablation) | FakeCLR with latent augmentation · FFHQ-100 | FID: 44.10 -> 41.62 | Improved |
| add forgetting factor of queue (ablation) | FakeCLR with latent augmentation · FFHQ-2K | FID: 10.65 -> 10.40 | Improved |
| add diversity-aware queue (ablation) | FakeCLR without it · FFHQ-100 | FID: 41.62 -> 42.56 | Hurt |
| add diversity-aware queue (ablation) | FakeCLR without it · FFHQ-2K | FID: 10.40 -> 9.90 | Improved |
Later papers that compared with this method
3 later accepted papers report FakeCLR as a baseline in their results tables; 2 of them report a better number on most of the results they share with it.
- CHAIN: Enhancing Generalization in Data-Efficient GANs via lipsCHitz continuity constrAIned Normalization (CVPR 2024 · better on 0 of 1 shared result)
- Rejection Sampling IMLE: Designing Priors for Better Few-Shot Image Synthesis (ECCV 2024 · better on 1 of 1 shared result)
- NICE: NoIse-modulated Consistency rEgularization for Data-Efficient GANs (NeurIPS 2023 · better on 2 of 2 shared results)
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.
- GenCo — GenCo: Generative Co-training for Generative Adversarial Networks with Limited Data (AAAI 2022) · better on 4 of 5 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
- CLR-GAN: Improving GANs Stability and Quality via Consistent Latent Representation and Reconstruction (ECCV 2024 · similarity 0.89)
- InfoGAN-CR and ModelCentrality: Self-supervised Model Training and Selection for Disentangling GANs (ICML 2020 · similarity 0.89)
- Training GANs with Stronger Augmentations via Contrastive Discriminator (ICLR 2021 · similarity 0.88)
- Consistency-GAN: Training GANs with Consistency Model (AAAI 2024 · similarity 0.88)
- Alleviation of Gradient Exploding in GANs: Fake Can Be Real (CVPR 2020 · similarity 0.88)
- Improving the Training of Data-Efficient GANs via Quality Aware Dynamic Discriminator Rejection Sampling (CVPR 2025 · similarity 0.88)
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.
