Research map / Image generation / GANs
Bridging Maximum Likelihood and Adversarial Learning via α-Divergence
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Use α-Bridge instead of vanilla DCGAN | CIFAR10 image generation · CIFAR10 | Inception score: 6.558 -> 7.225 | Helped |
| Use α-Bridge instead of vanilla DCGAN | CIFAR10 image generation · CIFAR10 | FID: 33.707 -> 28.083 | Helped |
| Use α-Bridge instead of FR, FRw, or FRw3 | Generative model with limited capacity · 1-D Gaussian mixture; means (6.0, -1.0, 1.0), standard deviations (0.3, 0.3, 0.3) | successful matches: 0/1/0 -> 10 | Helped |
| Use α-Bridge instead of FR | Expressive generative model · 1-D Gaussian mixture; means (6.0, -1.0, 1.0), standard deviations (1.0, 0.1, 0.1) | successful matches: 9 -> 10 | Helped |
| Use α-Bridge instead of FRw | Expressive generative model · 1-D Gaussian mixture; means (6.0, -1.0, 1.0), standard deviations (1.0, 0.1, 0.1) | successful matches: 4 -> 10 | Helped |
| Use α-Bridge instead of FRw3 | Expressive generative model · 1-D Gaussian mixture; means (6.0, -1.0, 1.0), standard deviations (1.0, 0.1, 0.1) | successful matches: 5 -> 10 | Helped |
| Vary α and γ schedules to sigmoid or linear-sigmoid schedules (ablation) | α-Bridge on 25-Gaussians · 25-Gaussians | Inception score and estimated log-likelihood | Mixed |
| Skip ML pretraining and begin with adversarial training (ablation) | α-Bridge · 25-Gaussians and MNIST | information retention | Hurt |
Closest papers by meaning
- Flow-GAN: Combining Maximum Likelihood and Adversarial Learning in Generative Models (AAAI 2018 · similarity 0.90)
- Adaptive Density Estimation for Generative Models (NeurIPS 2019 · similarity 0.89)
- Are GANs overkill for NLP? (NeurIPS 2022 · similarity 0.88)
- Improving Adversarial Energy-Based Model via Diffusion Process (ICML 2024 · similarity 0.87)
- Unified Generation, Reconstruction, and Representation: Generalized Diffusion with Adaptive Latent Encoding-Decoding (ICML 2024 · similarity 0.86)
- Direct Discriminative Optimization: Your Likelihood-Based Visual Generative Model is Secretly a GAN Discriminator (ICML 2025 · similarity 0.86)
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.