# CoRe-SAM3: Conditional Semantic--Visual Reconciliation for SAM3 Crack Segmentation

arXiv: https://arxiv.org/abs/2609.05816 · HTML: https://mlsanity.com/paper/2609.05816/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 22 of its 40 references that are in the map.
- First posted on arXiv 5 Sep 2026.

## Builds on (most cited first, 22 of 22)

- [Segment Anything](https://mlsanity.com/paper/2304.02643/) (2023)
- [Visual Prompt Tuning](https://mlsanity.com/paper/2203.12119/) (2022)
- [AdaptFormer: Adapting Vision Transformers for Scalable Visual Recognition](https://mlsanity.com/paper/2205.13535/) (2022)
- [Segment Everything Everywhere All at Once](https://mlsanity.com/paper/2304.06718/) (2023)
- [Segment Anything in High Quality](https://mlsanity.com/paper/2306.01567/) (2023)
- [clDice -- A Novel Topology-Preserving Loss Function for Tubular Structure Segmentation](https://mlsanity.com/paper/2003.07311/) (2020)
- [Scaling & Shifting Your Features: A New Baseline for Efficient Model Tuning](https://mlsanity.com/paper/2210.08823/) (2022)
- [Personalize Segment Anything Model with One Shot](https://mlsanity.com/paper/2305.03048/) (2023)
- [Semantic-SAM: Segment and Recognize Anything at Any Granularity](https://mlsanity.com/paper/2307.04767/) (2023)
- [Matcher: Segment Anything with One Shot Using All-Purpose Feature Matching](https://mlsanity.com/paper/2305.13310/) (2023)
- [FacT: Factor-Tuning for Lightweight Adaptation on Vision Transformer](https://mlsanity.com/paper/2212.03145/) (2022)
- [Convolution Meets LoRA: Parameter Efficient Finetuning for Segment Anything Model](https://mlsanity.com/paper/2401.17868/) (2024)
- [CAT-SAM: Conditional Tuning for Few-Shot Adaptation of Segment Anything Model](https://mlsanity.com/paper/2402.03631/) (2024)
- [CrackSeg9k: A Collection and Benchmark for Crack Segmentation Datasets and Frameworks](https://mlsanity.com/paper/2208.13054/) (2022)
- [SCSegamba: Lightweight Structure-Aware Vision Mamba for Crack Segmentation in Structures](https://mlsanity.com/paper/2503.01113/) (2025)
- [A Few-Shot Attention Recurrent Residual U-Net for Crack Segmentation](https://mlsanity.com/paper/2303.01582/) (2023)
- [SAM3-Adapter: Efficient Adaptation of Segment Anything 3 for Camouflage Object Segmentation, Shadow Detection, and Medical Image Segmentation](https://mlsanity.com/paper/2511.19425/) (2025)
- [Topology-aware Mamba for Crack Segmentation in Structures](https://mlsanity.com/paper/2410.19894/) (2024)
- [Multi-scale Contrastive Adaptor Learning for Segmenting Anything in Underperformed Scenes](https://mlsanity.com/paper/2408.05936/) (2024)
- [SAM3-UNet: Simplified Adaptation of Segment Anything Model 3](https://mlsanity.com/paper/2512.01789/) (2025)
- [MixerCSeg: An Efficient Mixer Architecture for Crack Segmentation via Decoupled Mamba Attention](https://mlsanity.com/paper/2603.01361/) (2026)
- [Rethinking Efficient Crack Segmentation with Task-Aligned Structural-Directional Modeling](https://mlsanity.com/paper/2605.31048/) (2026)

## Abstract (arXiv)

Crack segmentation requires a model to recognize target semantics while accurately recovering thin, low-contrast, and topologically continuous local structures. Although SAM3 provides strong open-concept segmentation, its direct application to the crack domain still misses weak cracks, activates crack-like background regions, and produces local boundary errors. We first diagnose the functional differences between the internal prompt-conditioned semantic representation and native visual representation of SAM3 on five crack datasets. The results show that the semantic representation already carries most task information for crack prediction, whereas the utility of the visual representation depends on the current semantic state. Directly combining the two representations does not yield consistent gains. Based on this finding, we propose Conditional Semantic--Visual Reconciliation, termed CoRe. CoRe retains semantic prediction as the primary decision path, applies lightweight semantic calibration to adjust the target-domain decision mapping, and uses spatially aligned native visual evidence to generate a zero-initialized, bounded, and regularized conditional residual that selectively corrects existing predictions. Across five domains, CoRe-SAM3 improves the average Crack IoU from 62.34% to 70.47% and clDice from 81.98% to 89.24%, while introducing only 18.914 K trainable parameters. Prediction-transition analysis further shows that CoRe corrects an average of 34.38% of native errors, with a damage rate of only 0.23% on pixels correctly classified by native SAM3. These results demonstrate that constrained prediction correction based on the functional differences between internal representations provides an effective and parameter-efficient target-domain adaptation strategy for vision foundation models with strong task-specific semantic priors.