Research map / Image generation / Image restoration
Single Image Reflection Removal Beyond Linearity
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 |
|---|---|---|---|
| use non-linear alpha-mask synthesis and train RmNet | single-image reflection removal · Zhang et al. [31] test | PSNR: 20.744 -> 21.283 | Helped |
| use non-linear alpha-mask synthesis and train RmNet | single-image reflection removal · SIR2 test | PSNR: 24.271 -> 23.707 | Hurt |
| use non-linear alpha-mask synthesis and train RmNet | single-image reflection removal · Synthetic all types | PSNR: 19.777 -> 21.656 | Helped |
| add a reconstruction loss that recombines predicted layers with the predicted mask (ablation) | RmNet · Zhang et al. [31] test | PSNR: 19.029 -> 21.283 | Helped |
| add a reconstruction loss that recombines predicted layers with the predicted mask (ablation) | RmNet · SIR2 test | PSNR: 21.011 -> 23.707 | Helped |
| add a reconstruction loss that recombines predicted layers with the predicted mask (ablation) | RmNet · Synthetic all types | PSNR: 20.274 -> 21.656 | Helped |
| add a gradient loss on the transmission layer (ablation) | RmNet · Zhang et al. [31] test | PSNR: 19.303 -> 21.283 | Helped |
| add a gradient loss on the transmission layer (ablation) | RmNet · SIR2 test | PSNR: 21.276 -> 23.707 | Helped |
| add a gradient loss on the transmission layer (ablation) | RmNet · Synthetic all types | PSNR: 20.229 -> 21.656 | Helped |
| fine-tune CEILNet using the diverse synthetic data | CEILNet · Zhang et al. [31] test | PSNR: 18.555 -> 16.753 | Hurt |
| fine-tune Zhang et al. [31] using the diverse synthetic data | Zhang et al. [31] · SIR2 test | PSNR: 24.271 -> 20.265 | Hurt |
| train using the complete pixel, gradient, and reconstruction objective instead of pixel loss only (ablation) | RmNet · Synthetic all types | PSNR: 19.413 -> 21.656 | Helped |
Closest papers by meaning
- Single Image Reflection Removal Through Cascaded Refinement (CVPR 2020 · similarity 0.90)
- FIRM: Flexible Interactive Reflection ReMoval (AAAI 2025 · similarity 0.89)
- Revisiting Single Image Reflection Removal In the Wild (CVPR 2024 · similarity 0.89)
- Removing Reflections from RAW Photos (CVPR 2025 · similarity 0.87)
- Flash-Split: 2D Reflection Removal with Flash Cues and Latent Diffusion Separation (CVPR 2025 · similarity 0.87)
- PolarFree: Polarization-based Reflection-Free Imaging (CVPR 2025 · 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.