Research map / Probabilistic and scientific modeling / Variational and generative sampling
Structured Output Learning with Conditional Generative Flows
You Lu, Bert Huang
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 c-Glow instead of DVN | Binary segmentation · Weizmann Horse, 32 × 32 | IOU: 0.840 -> 0.812 | Hurt |
| Use c-Glow instead of FCN | Binary segmentation · Weizmann Horse, 64 × 64 | IOU: 0.701 -> 0.852 | Improved |
| Use c-Glow instead of FCN | Binary segmentation · Weizmann Horse, 128 × 128 | IOU: 0.795 -> 0.858 | Improved |
| Use c-Glow instead of DVN | Multi-class segmentation · LFW, 32 × 32 | SPA: 0.924 -> 0.914 | Hurt |
| Use c-Glow instead of FCN | Multi-class segmentation · LFW, 64 × 64 | SPA: 0.792 -> 0.931 | Improved |
| Use c-Glow instead of FCN | Multi-class segmentation · LFW, 128 × 128 | SPA: 0.951 -> 0.945 | Hurt |
| Use c-Glow instead of DnCNN | Color image denoising · BSDS500 (σ = 25) | PSNR: 28.53 -> 27.61 | Hurt |
| Use c-Glow instead of BM3D | Color image denoising · BSDS500 (σ = 25) | PSNR: 28.21 -> 27.61 | Hurt |
| Use c-Glow instead of ProximalNet | Depth refinement · Seven Scenes | PSNR: 36.31 -> 36.53 | Improved |
| Use c-Glow instead of DCGANi | Image inpainting · CelebA, central block mask | PSNR: 22.73 -> 24.88 | Improved |
| Add Poisson blending to DCGANi | DCGANi inpainting · CelebA, central block mask | PSNR: 22.73 -> 23.65 | Improved |
Closest papers by meaning
- Glow: Generative Flow with Invertible 1x1 Convolutions (NeurIPS 2018 · similarity 0.86)
- A Likelihood Based Approach to Distribution Regression Using Conditional Deep Generative Models (ICML 2025 · similarity 0.84)
- Structured Generative Adversarial Networks (NeurIPS 2017 · similarity 0.84)
- Active Learning of Conditional Generative Models via the Transport Neural Tangent Kernel (NeurIPS 2026 · similarity 0.84)
- Don't Drop Your Samples! Coherence-Aware Training Benefits Conditional Diffusion (CVPR 2024 · similarity 0.84)
- CAR-Flow: Condition-Aware Reparameterization Aligns Source and Target for Better Flow Matching (NeurIPS 2025 · similarity 0.83)
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.
