Research map / Image generation / Image restoration
Sea-Thru: A Method for Removing Water From Underwater Images
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 the revised underwater model with distinct backscatter and distance-dependent attenuation coefficients instead of the former model with a shared coefficient (ablation) | Underwater color recovery · D1–D5 average | RGB angular error (degrees): 21.77 -> 6.33 | Helped |
| Use the revised underwater model with distinct backscatter and distance-dependent attenuation coefficients instead of the former model with a shared coefficient (ablation) | Underwater color recovery · Stereo dataset [7] average | RGB angular error (degrees): 12.04 -> 4.94 | Helped |
| Estimate distance-dependent attenuation with Sea-thru rather than use the direct-signal illuminant correction without explicitly modeling attenuation (ablation) | Revised-model color recovery · D1–D5 average | RGB angular error (degrees): 4.13 -> 6.33 | Hurt |
| Estimate distance-dependent attenuation with Sea-thru rather than use the direct-signal illuminant correction without explicitly modeling attenuation (ablation) | Revised-model color recovery · Stereo dataset [7] average | RGB angular error (degrees): 4.46 -> 4.94 | Hurt |
| Use Sea-thru instead of DCP-based backscatter estimation with the former model | Underwater color recovery · D1–D5 average | RGB angular error (degrees): 14.38 -> 6.33 | Helped |
| Use Sea-thru instead of DCP-based backscatter estimation with the former model | Underwater color recovery · Stereo dataset [7] average | RGB angular error (degrees): 10.03 -> 4.94 | Helped |
| Use Sea-thru instead of global contrast stretch | Underwater color recovery · D1–D5 average | RGB angular error (degrees): 12.49 -> 6.33 | Helped |
| Use Sea-thru instead of global contrast stretch | Underwater color recovery · Stereo dataset [7] average | RGB angular error (degrees): 6.83 -> 4.94 | Helped |
Closest papers by meaning
- WaterDiffusion: Learning a Prior-involved Unrolling Diffusion for Joint Underwater Saliency Detection and Visual Restoration (AAAI 2025 · similarity 0.87)
- Atlantis: Enabling Underwater Depth Estimation with Stable Diffusion (CVPR 2024 · similarity 0.86)
- NAUTILUS: A Large Multimodal Model for Underwater Scene Understanding (NeurIPS 2025 · similarity 0.85)
- Sea-ing in Low-light (CVPR 2025 · similarity 0.85)
- Looking Into the Water by Unsupervised Learning of the Surface Shape (NeurIPS 2025 · similarity 0.85)
- Neural Underwater Scene Representation (CVPR 2024 · 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.