Research map / Video / Video segmentation and tracking
Self-Supervised Video Desmoking for Laparoscopic Surgery
Renlong Wu, Zhilu Zhang, Shuohao Zhang, Longfei Gou, Haobin Chen, Lei Zhang, Hao Chen, Wangmeng Zuo
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 deformation-based reconstruction loss using PS frame supervision (ablation) | SelfSVD · LSVD | PSNR: 22.40 -> 24.23 | Improved |
| align output to PS frame via optical flow instead of naive l1 (ablation) | SelfSVD · LSVD | PSNR: 22.67 -> 24.23 | Improved |
| use flow-based alignment instead of CoBi loss (ablation) | SelfSVD · LSVD | PSNR: 22.82 -> 24.23 | Improved |
| warp output to PS frame instead of PS frame to output (ablation) | SelfSVD · LSVD | PSNR: 24.12 -> 24.23 | Improved |
| add PS frame as reference input (ablation) | SelfSVD · LSVD | PSNR: 23.56 -> 24.23 | Improved |
| use PS frame as Ref instead of first smoky frame (ablation) | SelfSVD · LSVD | PSNR: 23.95 -> 24.23 | Improved |
| add both masking strategy and regularization to avoid trivial solutions (ablation) | SelfSVD · LSVD | visual quality | Improved |
| fine-tune with enhanced PS frame supervision (SelfSVD*) | SelfSVD · LSVD | PSNR: 24.23 -> 24.58 | Improved |
| use video methods instead of single-image methods | desmoking models · LSVD | PSNR | Improved |
| use PSD and DCP dehazing priors | unsupervised image methods · LSVD | FADE vs NIQE/PI: FADE 0.1715 (PSD) | Mixed |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- BasicVSR++ — BasicVSR: The Search for Essential Components in Video Super-Resolution and Beyond (CVPR 2021) · better on 3 of 3 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Rethinking Surgical Smoke: A Smoke-Type-Aware Laparoscopic Video Desmoking Method and Dataset (AAAI 2026 · similarity 0.93)
- Benchmarking Endoscopic Surgical Image Restoration and Beyond (CVPR 2026 · similarity 0.82)
- SelfHVD: Self-Supervised Handheld Video Deblurring (CVPR 2026 · similarity 0.82)
- MedCache: Training-Free Spatially Aware Caching for Accelerated Medical Video Generation (NeurIPS 2026 · similarity 0.82)
- SelfPromer: Self-Prompt Dehazing Transformers with Depth-Consistency (AAAI 2024 · similarity 0.81)
- Intraoperative 2D/3D Image Registration via Differentiable X-ray Rendering (CVPR 2024 · similarity 0.80)
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.
