Research map / Detection and segmentation / CNN features
SCRWKV: Ultra-Compact Structure-Calibrated Vision-RWKV for Topological Crack Segmentation
Hanxu Zhang, Chen Jia, Hui Liu, Xu Cheng, Fan Shi, Shengyong Chen
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 GBST and DSCD to AMCM-only model (ablation) | SCRWKV · TUT | F1: 0.8359 -> 0.8428 | Improved |
| add AMCM to GBST+DSCD (ablation) | SCRWKV · TUT | F1: 0.8295 -> 0.8428 | Improved |
| add GBST to AMCM+DSCD (ablation) | SCRWKV · TUT | F1: 0.8381 -> 0.8428 | Improved |
| add DSCD to AMCM+GBST (ablation) | SCRWKV · TUT | F1: 0.8340 -> 0.8428 | Improved |
| use CSHF decoder instead of MFS head (ablation) | SCRWKV · TUT | F1: 0.8374 -> 0.8428 | Improved |
| use GBST instead of Q-Shift (ablation) | SCRWKV · TUT | F1: 0.8381 -> 0.8428 | Improved |
| use GBST instead of DiagSnake (ablation) | SCRWKV · TUT | F1: 0.8397 -> 0.8428 | Improved |
| use 4 SCIU layers instead of 2 (ablation) | SCRWKV · TUT | F1: 0.8314 -> 0.8428 | Improved |
| increase SCIU layers from 4 to 16 (ablation) | SCRWKV · TUT | mIoU: 0.8512 -> 0.8494 | Hurt |
| increase SCIU layers from 4 to 8 (ablation) | SCRWKV · TUT | F1: 0.8428 -> 0.8386 | Hurt |
| increase patch size from 4 to 32 (ablation) | SCRWKV · TUT | F1: 0.8428 -> 0.7390 | Hurt |
| increase patch size from 4 to 8 (ablation) | SCRWKV · TUT | F1: 0.8428 -> 0.8342 | Hurt |
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.
- SCSegamba — SCSegamba: Lightweight Structure-Aware Vision Mamba for Crack Segmentation in Structures (CVPR 2025) · better on 7 of 9 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
- CrackSSM: Reviving SSMs for Crack Segmentation via Dynamic Scanning (CVPR 2026 · similarity 0.88)
- SCSegamba: Lightweight Structure-Aware Vision Mamba for Crack Segmentation in Structures (CVPR 2025 · similarity 0.88)
- Prism: Harmonizing Missing Modalities via Implicit Structural Alignment on Lightweight Pulse RWKV for Multimodal Crack Segmentation (NeurIPS 2026 · similarity 0.88)
- MixerCSeg: An Efficient Mixer Architecture for Crack Segmentation via Decoupled Mamba Attention (CVPR 2026 · similarity 0.87)
- Cracks in the Foundation: A Civil Infrastructure Dataset to Challenge Vision Foundation Models (NeurIPS 2026 · similarity 0.84)
- Mind Marginal Non-Crack Regions: Clustering-Inspired Representation Learning for Crack Segmentation (CVPR 2024 · similarity 0.83)
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.
