Research map / Image generation / Autoregressive and latent generation
Plug & Play Generative Networks: Conditional Iterative Generation of Images in Latent Space
Check 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 |
|---|---|---|---|
| Add a learned latent prior to DGN-AM (ablation) | DGN-AM, producing Noiseless Joint PPGN-h · ImageNet classes | sample diversity | Helped |
| Add a learned latent prior to DGN-AM (ablation) | DGN-AM, producing Noiseless Joint PPGN-h · ImageNet classes | sample quality | Helped |
| Model the latent prior through the generator and encoder composition G(E(.)) | PPGN-h · ImageNet classes | sample quality | Helped |
| Replace the separately trained MLP prior with a joint generator-encoder prior | PPGN-h · ImageNet classes | mixing speed | Helped |
| Remove training noise and train with image reconstruction, pool5 feature matching, and adversarial losses (ablation) | Joint PPGN-h · ImageNet classes | sample quality | Helped |
| Add adversarial loss (ablation) | Noiseless Joint PPGN-h · ImageNet classes | sample quality | Helped |
| Use pool5 instead of fc6 feature matching (ablation) | Noiseless Joint PPGN-h · ImageNet classes | sample quality | Helped |
| Use fc6 feature matching instead of pool5 (ablation) | Noiseless Joint PPGN-h · ImageNet classes | sample quality | Hurt |
| Add noise during training (ablation) | Joint PPGN-h · ImageNet classes | sample quality | No clear effect |
| Increase sampling noise (ablation) | Noiseless Joint PPGN-h · ImageNet classes | sample quality | Hurt |
| Increase the prior multiplier too much (ablation) | Noiseless Joint PPGN-h · ImageNet classes | class specificity | Hurt |
| Replace the class classifier condition with an MIT Places classifier | PPGN · MIT Places classes | sample quality | Helped |
Closest papers by meaning
- Boosting Generative Image Modeling via Joint Image-Feature Synthesis (NeurIPS 2025 · similarity 0.89)
- Discovering Latent Graphs with GFlowNets for Diverse Conditional Image Generation (NeurIPS 2025 · similarity 0.88)
- Latent Zoning Network: A Unified Principle for Generative Modeling, Representation Learning, and Classification (NeurIPS 2025 · similarity 0.88)
- Attack Deterministic Conditional Image Generative Models for Diverse and Controllable Generation (AAAI 2024 · similarity 0.88)
- Direct Discriminative Optimization: Your Likelihood-Based Visual Generative Model is Secretly a GAN Discriminator (ICML 2025 · similarity 0.88)
- BiGR: Harnessing Binary Latent Codes for Image Generation and Improved Visual Representation Capabilities (ICLR 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.