Research map / Image generation / GANs
Improving Generative Adversarial Networks With Local Coordinate Coding
Jiezhang Cao, Yong Fang Guo, Qingyao Wu, Chunhua Shen, Junzhou Huang, Mingkui Tan
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 LCC sampling (q=2) to DCGAN (ablation) | DCGAN · Oxford-102 | FID: 182.2 -> 61.9 | Improved |
| use higher-order term (q=3) instead of q=2 in LCC (ablation) | LCCGAN · Oxford-102 | FID: 61.9 -> 57.7 | Improved |
| add LCC (q=3) to WGAN-GP (ablation) | WGAN-GP · Oxford-102 | FID: 160.4 -> 145.9 | Improved |
| add LCC (q=3) to StackGAN-v1 (ablation) | StackGAN-v1 · Oxford-102 | FID: 178.8 -> 168.2 | Improved |
| add LCC (q=3) to StackGAN-v2 (ablation) | StackGAN-v2 · Oxford-102 | FID: 27.0 -> 22.2 | Improved |
| add LCC (q=3) to Progressive GAN (ablation) | Progressive GAN · Oxford-102 | FID: 114.5 -> 109.6 | Improved |
| add LCC (q=2) to StackGAN-v2 (ablation) | StackGAN-v2 · Oxford-102 | IS: 3.087 -> 3.088 | No clear effect |
| increase L_nu from 0.0001 to 0.001 (ablation) | LCCGAN++ · Oxford-102 (d=30) | FID: 57.7 -> 205.9 | Hurt |
| set L_h to 1 instead of 0.0001 (ablation) | LCCGAN++ · Oxford-102 (d=30) | FID: 247.9 -> 57.7 | Improved |
| increase number of bases M from 64 to 128 (ablation) | LCCGAN++ · Oxford-102 (d=30) | FID: 66.3 -> 57.7 | Improved |
| increase number of bases M from 128 to 256 (ablation) | LCCGAN++ · Oxford-102 (d=30) | FID: 57.7 -> 76.8 | Hurt |
| use LCCGAN++ with d=5 instead of d=100 DCGAN | LCCGAN++ vs DCGAN · Oxford-102 | FID: 182.2 -> 71.0 | Improved |
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.
- StackGAN — StackGAN: Text to Photo-Realistic Image Synthesis With Stacked Generative Adversarial Networks (ICCV 2017) · better on 2 of 2 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
- Adversarial Learning with Local Coordinate Coding (ICML 2018 · similarity 0.97)
- Improving Generative Adversarial Networks via Adversarial Learning in Latent Space (NeurIPS 2022 · similarity 0.89)
- Global Versus Localized Generative Adversarial Nets (CVPR 2018 · similarity 0.89)
- CLR-GAN: Improving GANs Stability and Quality via Consistent Latent Representation and Reconstruction (ECCV 2024 · similarity 0.89)
- Improving GAN with Neighbors Embedding and Gradient Matching (AAAI 2019 · similarity 0.88)
- MCGAN: Enhancing GAN Training with Regression-Based Generator Loss (AAAI 2025 · similarity 0.87)
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.
