Research map / Image generation / Text-to-image diffusion
DPGF-Net: Dual-Prior Guided Fusion Network for Joint Assessment of Perceptual Quality and Semantic Consistency in AI-Generated Images
Tao Li, Xingran Liao, Mingliang Zhou
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 both TCPGA and FIM modules (ablation) | DPGF-Net · AIGCIQA2023 alignment | SRCC: 0.7964 -> 0.8278 | Improved |
| add both TCPGA and FIM modules (ablation) | DPGF-Net · AIGCIQA2023 quality | SRCC: 0.8647 -> 0.8806 | Improved |
| add TCPGA local text-conditioned aggregation (ablation) | DPGF-Net · AIGCIQA2023 quality | SRCC: 0.8744 -> 0.8806 | Improved |
| add FIM global modulation path (ablation) | DPGF-Net · AIGCIQA2023 alignment | SRCC: 0.8015 -> 0.8278 | Improved |
| add TCPGA to a model that already has FIM (ablation) | DPGF-Net · AIGCIQA2023 alignment | SRCC: 0.8015 -> 0.8278 | Improved |
| increase sampled patches from 4 to 12 (ablation) | DPGF-Net · AIGCIQA2023 average | average correlation: 0.8490 -> 0.8544 | Improved |
| increase sampled patches from 12 to 14 (ablation) | DPGF-Net · AIGCIQA2023 average | average correlation: 0.8544 -> 0.8510 | Hurt |
| use dual-prior fusion instead of IPCE in cross-dataset transfer | DPGF-Net · AIGCIQA2023 to AGIQA-3K, quality | PLCC: 0.7519 -> 0.8235 | Improved |
| use DPGF-Net instead of IPCE on AGIQA-3K alignment PLCC | DPGF-Net · AGIQA-3K alignment | PLCC: 0.8725 -> 0.8698 | Hurt |
Closest papers by meaning
- Reasoning as Representation: Rethinking Visual Reinforcement Learning in Image Quality Assessment (ICLR 2026 · similarity 0.86)
- VisualQuality-R1: Reasoning-Induced Image Quality Assessment via Reinforcement Learning to Rank (NeurIPS 2025 · similarity 0.86)
- Q-Insight: Understanding Image Quality via Visual Reinforcement Learning (NeurIPS 2025 · similarity 0.86)
- AGFSync: Leveraging AI-Generated Feedback for Preference Optimization in Text-to-Image Generation (AAAI 2025 · similarity 0.85)
- UFO: Chain-of-Evaluation for Omni-Condition Alignment in Multi-Modal Image Generation (ICML 2026 · similarity 0.85)
- Q-Probe: Scaling Image Quality Assessment to High Resolution via Context-Aware Agentic Probing (NeurIPS 2026 · similarity 0.85)
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.
