Research map / Probabilistic and scientific modeling / Variational and generative sampling
Conditional Generative Modeling via Learning the Latent Space
Sameera Ramasinghe, Kanchana Nisal Ranasinghe, Salman Khan, Nick Barnes, Stephen Gould
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 |
|---|---|---|---|
| Use latent-space optimization with momentum correction instead of a fixed-output conditional generator | Conditional Generative Modeling via Learning the Latent Space · Image completion, 15% corruption | PSNR: 20.50 -> 26.43 | Improved |
| Use latent-space optimization with momentum correction instead of a fixed-output conditional generator | Conditional Generative Modeling via Learning the Latent Space · Image completion, 25% corruption | LPIP: 0.147 -> 0.129 | Improved |
| Use latent-space optimization with momentum correction instead of a fixed-output conditional generator | Conditional Generative Modeling via Learning the Latent Space · Image completion, 10% corruption | LPIP: 0.045 -> 0.051 | Hurt |
| Add momentum correction ρ to latent updates (ablation) | Proposed method · Image completion, 15% corruption | PSNR: 23.22 -> 26.43 | Improved |
| Add momentum correction ρ to latent updates (ablation) | Proposed method · Image completion, 25% corruption | SSIM: 0.74 -> 0.77 | Improved |
| Set latent dimension to 16 rather than 5 (ablation) | Proposed method · Colorization, 128 × 128 inputs | LPIP: 1.05 -> 0.14 | Improved |
| Set latent dimension to 32 rather than 16 (ablation) | Proposed method · Colorization, 128 × 128 inputs | LPIP: 0.14 -> 0.12 | Improved |
| Increase latent dimension from 32 to 64 (ablation) | Proposed method · Colorization, 128 × 128 inputs | LPIP: 0.12 -> 0.27 | Hurt |
| Increase latent dimension from 64 to 128 (ablation) | Proposed method · Colorization, 128 × 128 inputs | LPIP: 0.27 -> 0.69 | Hurt |
| Increase latent dimension from 32 to 64 (ablation) | Proposed method · Colorization, 128 × 128 inputs | Diversity: 0.043 -> 0.048 | Improved |
Closest papers by meaning
- Latent Constraints: Learning to Generate Conditionally from Unconditional Generative Models (ICLR 2018 · similarity 0.89)
- Learning Multimodal Latent Generative Models with Energy-Based Prior (ECCV 2024 · similarity 0.89)
- Linear combinations of latents in generative models: subspaces and beyond (ICLR 2025 · similarity 0.87)
- A Likelihood Based Approach to Distribution Regression Using Conditional Deep Generative Models (ICML 2025 · similarity 0.87)
- A Theory for Conditional Generative Modeling on Multiple Data Sources (ICML 2025 · similarity 0.87)
- Stabilize the Latent Space for Image Autoregressive Modeling: A Unified Perspective (NeurIPS 2024 · 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.
