Research map / Image generation / Image restoration
Nighttime Flare Removal via Wavelet-Guided and Gated-Enhanced Spatial-Frequency Fusion Network
Yun Liu, Guang Yang, Tao Li, Weisi Lin
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 SFFN (spatial-frequency fusion) (ablation) | WGSF-Net · Flare7K synthetic test set | PSNR: 29.26 -> 30.91 | Improved |
| add GEAB (gated attention) (ablation) | WGSF-Net · Flare7K synthetic test set | PSNR: 29.51 -> 30.91 | Improved |
| add MWEB (multi-level wavelet) (ablation) | WGSF-Net · Flare7K real test set | LPIPS: 0.0419 -> 0.0383 | Improved |
| add ResBlock (ablation) | WGSF-Net · Flare7K real test set | LPIPS: 0.0397 -> 0.0383 | Improved |
| add WFB (ablation) | WGSF-Net · Flare7K real test set | PSNR: 28.25 -> 28.52 | Improved |
| add LGB (ablation) | WGSF-Net · Flare7K real test set | PSNR: 28.09 -> 28.52 | Improved |
| add Lvgg to L1 loss (ablation) | WGSF-Net · Flare7K real test set | LPIPS: 0.0491 -> 0.0415 | Improved |
| add Lfft to L1+Lvgg loss (ablation) | WGSF-Net · Flare7K synthetic test set | PSNR: 30.22 -> 30.01 | Hurt |
| add Lfft to L1+Lvgg loss (ablation) | WGSF-Net · Flare7K real test set | PSNR: 28.35 -> 28.46 | Improved |
| add Lrec to full loss (ablation) | WGSF-Net · Flare7K real test set | PSNR: 28.46 -> 28.52 | Improved |
| WGSF-Net vs best prior (LPFSformer) | Flare7K++ training · Flare7K real test set | PSNR: 28.24 -> 28.52 | Improved |
| WGSF-Net vs Dai (Flare7K) on synthetic LPIPS | Flare7K synthetic test set · Flare7K synthetic test set | LPIPS: 0.017 vs 0.018 | No clear effect |
Closest papers by meaning
- Flare7K++: Mixing Synthetic and Real Datasets for Nighttime Flare Removal and Beyond (TPAMI 2024 · similarity 0.90)
- Flare7K: A Phenomenological Nighttime Flare Removal Dataset (NeurIPS 2022 · similarity 0.90)
- SLCFormer: Spectral-Local Context Transformer with Physics-Grounded Flare Synthesis for Nighttime Flare Removal (AAAI 2026 · similarity 0.89)
- Language-Guided One-Step Diffusion Model for Nighttime Flare Removal (CVPR 2026 · similarity 0.87)
- Disentangle Nighttime Lens Flares: Self-supervised Generation-based Lens Flare Removal (AAAI 2025 · similarity 0.86)
- FlareX: A Physics-Informed Dataset for Lens Flare Removal via 2D Synthesis and 3D Rendering (NeurIPS 2025 · similarity 0.84)
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.
