Pith. sign in

REVIEW 3 major objections 4 minor 30 references

Toward Mask Annotation-Free Surgical Instrument Segmentation from Endoscopic Images Using Text-Prompted Segment Anything Model 3 (SAM3)

T0 review · 3 major / 4 minor · reviewed 2026-08-14 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read SAM3 with the generic prompt 'tool' plus a Qwen classifier gives instance-level surgical instrument segmentation without pixel-level masks or manual prompts.

desk verdict Useful empirical study of SAM3 text prompting for surgical instruments, but the mask-annotation-free instance segmentation claim collapses on the undisclosed Qwen label source. read the letter →

arxiv 2608.08844 v1 pith:3PSDDTOF submitted 2026-08-09 cs.CV

Nakul Poudel , Richard Simon , Cristian A. Linte This is my paper · ORCID
classification cs.CV
keywords surgicalinstrumentsegmentationSegmentAnythingModel3text-promptedvision-languagezero-shotmaskannotation-freeEndoVisQwenfine-tuning
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Surgical instrument segmentation, a building block for computer-assisted interventions, normally demands pixel-level masks or manual point or box prompts; this paper claims a route that needs neither. The proposal is a two-stage pipeline: prompt SAM3 with the generic word "tool" to obtain binary instrument masks, then fine-tune the Qwen vision-language model on those masked regions to assign each mask an instrument label. On EndoVis 2017 and 2018, the pipeline produces instance-level segmentation without ground-truth masks or spatial interaction, clearly outperforming SAM3 prompted with instrument names, although it still trails fully supervised systems. The paper argues this is a viable direction toward annotation-free surgical tool segmentation, which matters because annotation cost is the main scalability bottleneck in the field.

What carries the argument

The machinery is a two-stage localize-then-identify loop. Stage one treats SAM3 as a zero-shot text-prompted proposal generator: a fixed prompt "tool" yields candidate binary masks, filtered by confidence threshold $t=0.2$ and merged by an IoU threshold $m=0.02$. Stage two extracts each mask's region by element-wise multiplication with the RGB frame and feeds it to Qwen 2.5-7B, adapted with rank-8 LoRA on SAM3-generated masked regions, to output one of seven instrument classes or "No Tool". The load-bearing operation is the decoupling: the generic prompt "tool" avoids the domain gap that breaks instrument-name prompts, and the downstream classifier supplies the semantic label, so the pipeline never needs a spatial prompt or a pixel mask.

What would settle it

Run the Qwen fine-tuning script with every ground-truth mask file removed from the workspace. If the training set still assigns correct instrument labels and a working "No Tool" set to SAM3-generated masks, the mask-annotation-free claim holds; the moment label assignment reads a ground-truth mask, for instance by maximum-overlap matching, the central claim fails.

Watch

Extended reading notes

Core claim

The central claim is that text-prompted segmentation can work in surgery if the text prompt is category-agnostic rather than instrument-specific. SAM3, prompted with "tool" and then thresholded and mask-merged, generates binary masks; Qwen, fine-tuned with LoRA on the masked image regions, names each mask's instrument, and a "No Tool" class discards false-positive masks. The result is instance-level segmentation maps built from the input image and a fixed text string, with no manual prompt and no pixel-level ground truth. The paper reports binary Dice near 82-85 (comparable to a supervised UNet), instance Ch_IoU of 40.56 and 61.53 across the two EndoVis datasets, and an oracle-classifier analysis showing that with a perfect classifier the same masks would reach Ch_IoU of 61.49 and 79.39, close to some supervised methods.

Load-bearing premise

The framework is only genuinely mask annotation-free if every label used to fine-tune Qwen, including the "No Tool" category, can be produced from SAM3 masks and text alone, without ever consulting a ground-truth mask.

Editorial extensions

If this is right

  • Binary instrument detection reaches Dice of 82.10 on EndoVis 2017 and 84.93 on EndoVis 2018 with the generic "tool" prompt, essentially matching a fully supervised UNet without training a segmentation model.
  • Instance-level Ch_IoU improves from 4.23 to 40.56 on EndoVis 2017 and from 7.79 to 61.53 on EndoVis 2018 compared with prompting SAM3 by instrument name.
  • A classifier fine-tuned on SAM3-generated masks rather than ground-truth masks loses only modest accuracy (e.g., 72.33 vs 76.67 on EndoVis 2018) and transfers across datasets with nearly equivalent performance on shared instrument categories.
  • An oracle classifier raises Ch_IoU to 61.49 on EndoVis 2017 and 79.39 on EndoVis 2018, showing mask quality from SAM3 is not the dominant bottleneck; classification quality is.
  • Mask errors from the first stage, such as fragmentation, overlap, and missed detections, propagate into the classification stage and account for the main failure modes, so future gains depend on improving SAM3 mask quality or adding a refinement stage.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Editorial extension: the paper never says how each SAM3 mask gets its instrument label during fine-tuning; if the labels come from maximum-overlap matching to ground-truth masks, the honest claim shrinks to "no pixel-level mask supervision for the segmentation stage," not "no mask annotation at all."
  • Editorial extension: if "No Tool" examples are instead defined by low confidence or Qwen's own uncertainty, the same design becomes a self-training loop that could run on fully unlabelled surgical videos, an extension the paper allows but does not state.
  • Editorial extension: the oracle results imply that investing in better text-prompted or language-only classification is likely to raise instance IoU more than tuning SAM3's mask thresholds, because the oracle nearly closes the gap to supervised methods.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 4 minor

Summary. The paper presents a two-stage approach to surgical instrument segmentation from endoscopic images. Stage 1 uses the pretrained SAM3 model with a fixed text prompt "tool" to generate binary masks, and Stage 2 uses a Qwen 2.5-7B vision-language model fine-tuned on the resulting masked regions to assign each mask an instrument class or a "No Tool" label. The authors claim this pipeline achieves instance-level segmentation without pixel-level ground truth masks and without manual spatial interaction. Experiments on EndoVis 2017, EndoVis 2018, and GraSP report binary IoU/Dice near or above a vanilla UNet baseline, instance-level metrics below fully supervised methods but far above a SAM3 instrument-name baseline, and an oracle analysis that isolates classification errors from mask generation errors.

Significance. The paper's strongest empirical contribution is the observation that a generic natural-language prompt ("tool") substantially outperforms instrument-name prompts for SAM3 on endoscopic images (Table 2), and the clean separation of mask localization from classification through the oracle analysis (Tables 3-4). The two-stage design is simple and the evaluation on three public benchmarks is clearly reported. If the mask-annotation-free claim were supported, this would be a meaningful step toward reducing annotation cost for surgical instrument segmentation. However, the central claim is currently unsupported because the paper does not explain how the training labels for Qwen are obtained without using mask-level ground truth.

major comments (3)
  1. [§3.3, §4.2] The label source for Qwen fine-tuning is not described. Section 3.3 states that Qwen is fine-tuned on SAM3-predicted masked regions and asserts that this "maintains a fully mask annotation-free training pipeline," but it never specifies how each masked region is assigned its class label during training. A supervised fine-tuning setup requires labels for every training region, including the "No Tool" category introduced in Section 4.2. Please state the exact labeling protocol, including the source of labels, the treatment of masks that overlap multiple instruments, and how "No Tool" negatives are identified. If any step uses pixel-level ground truth masks (for example, by matching each SAM3 mask to the ground truth with maximum IoU) or manual review of candidate masks, the abstract and Section 3.3 must be revised because the stated claim would be false.
  2. [§4.2] The construction of the "No Tool" category implies access to mask-level supervision. Section 4.2 introduces this category because "SAM3 can mis-segment tissue regions as instruments." Constructing it requires knowing, for each SAM3 mask, whether it is a true instrument or a false positive; this knowledge is exactly the pixel-level annotation the paper claims to avoid. The manuscript does not describe any mask-free mechanism for obtaining these labels. This is not a minor implementation detail: the "No Tool" filter is part of the proposed pipeline, and its presence in the training data implies either ground truth mask overlap, manual inspection, or an undisclosed heuristic. Please provide the mechanism and, ideally, an ablation without the "No Tool" category to measure how much of the reported performance depends on it.
  3. [§4.2, Table 1] The evaluation of Qwen on ground-truth masked regions replaces every "No Tool" prediction with a randomly chosen valid instrument class and averages over five trials. This makes the reported accuracy and macro F1 numbers in Table 1 dependent on an arbitrary stochastic procedure and does not reflect how the classifier is actually used in the full pipeline. Please report the confusion matrix including the "No Tool" column, and either treat "No Tool" as background in the downstream segmentation evaluation or justify the random-replacement procedure with a sensitivity analysis.
minor comments (4)
  1. [§4.3] The direct SAM3 baseline is given the ground-truth list of instrument names present in each frame to construct prompts, so the comparison is not strictly apples-to-apples with the proposed pipeline; please state this asymmetry explicitly when interpreting the large improvement over the baseline.
  2. [§4.4] The confidence threshold t=0.2 and the IoU merging threshold m=0.02 are fixed across datasets, and Figure 3 only investigates t on EndoVis 2018. Please clarify how these values were selected and whether the qualitative conclusions are robust to m.
  3. [Table 2] The UNet row on EndoVis 2018 is blank; adding the value would help the reader judge whether the binary segmentation result transfers across datasets.
  4. [§3.3, §5] When Qwen predicts "No Tool" during full inference, the paper does not state whether the mask is discarded, assigned to background, or handled differently in constructing the final segmentation map S; please specify this in the method description.

Circularity Check

1 steps flagged · score 6.0 of 10

The 'No Tool' fine-tuning labels silently reintroduce the ground-truth mask annotations the paper claims to avoid; the central mask-annotation-free claim is circular by construction.

  1. self definitional [Section 4.2 (Qwen Fine-tuning and Testing Setup); Section 3.3 (Tool Classification with Qwen)]
    "We fine-tuned the Qwen model on the SAM3-masked regions. As SAM3 can mis-segment tissue regions as instruments, we introduced an additional 'No Tool' category during fine-tuning. ... Importantly, fine-tuning Qwen with masked regions derived from SAM3-predicted masks, rather than ground truth masks, maintains a fully mask annotation-free training pipeline."

    Fine-tuning Qwen requires labeled training examples: each SAM3 mask must receive an instrument class or the 'No Tool' label. The 'No Tool' category is defined as SAM3 false positives on tissue, and identifying which SAM3 masks are false positives cannot be done from unlabeled SAM3 masks alone; it requires comparing them to ground-truth mask annotations or performing manual pixel-level review. Thus the same mask-level supervision the paper claims to eliminate is required to construct the classifier's training set. The assertion that training on 'SAM3-predicted masks' maintains a 'mask annotation-free' pipeline is therefore true only by ignoring the label source, and if the labels come from ground-truth masks the central claim fails by construction.

full rationale

The paper's genuinely non-circular component is the binary segmentation stage: SAM3 with the fixed 'tool' prompt is used zero-shot and evaluated on held-out test frames, so the binary IoU/Dice results are independent of the claimed derivation. The instance-level stage, however, has a hidden circular dependence. Qwen is fine-tuned on SAM3 masked regions, but the paper never specifies how each region receives its class label. The introduction of the 'No Tool' category during fine-tuning proves that some external signal identifies SAM3 false positives, and that signal is exactly the pixel-level annotation or manual inspection the paper claims to avoid. If ground-truth masks are used to create these labels, then the 'mask annotation-free training pipeline' is circular by construction: the avoided annotations are the source of the supervision. If manual review is used instead, the 'without manual interaction' claim fails. This is not a minor self-citation issue; it affects the central novelty. The self-citation to [19] for choosing Qwen is ordinary and not load-bearing, since the model is subsequently fine-tuned and evaluated on held-out data. Threshold tuning (t=0.2, m=0.02) is a benchmark-fitting concern rather than circularity. Because the binary result stands on its own but the instance-level claim reduces to an undisclosed annotation input, a partial circularity score of 6 is appropriate.

Assumptions & free parameters 5 free parameters · 5 assumptions · 0 invented entities

The reported pipeline depends on two large pretrained models, five hand-set hyperparameters, and an undisclosed labeling procedure. The absence of a described label source for Qwen fine-tuning is the single most important ledger item, because it bears directly on the central 'no ground truth masks' claim.

free parameters (5)
  • Generic text prompt 'tool' = 'tool'
    Hand-selected after observing that instrument-specific names produce poor SAM3 masks; no systematic prompt search is reported (Sections 3.3, 4.4).
  • Mask confidence threshold t = 0.2
    Set in Section 4.4; no ablation or sensitivity analysis is provided, so its effect on the central results is unknown.
  • IoU merging threshold m = 0.02
    Set in Section 4.4; used to merge overlapping SAM3 masks, no ablation shown.
  • Qwen fine-tuning setup = Rank-8 LoRA, 5 epochs, batch size 4, lr 1e-4, gradient accumulation 4
    Reported in Section 4.4; these hyperparameters are chosen without a search or sensitivity study.
  • Qwen LoRA adapter weights = learned on training split of each dataset
    The trained classifier weights are the fitted component of the second stage; they are learned, not independently verified.
assumptions (5)
  • domain assumption SAM3 can generate useful binary masks for surgical instruments from the generic text prompt 'tool'.
    This is the core of stage one; Table 2 supports it on EndoVis data, but it depends on the pretrained model's concept space and cannot be guaranteed outside these distributions.
  • domain assumption Qwen2.5-VL can classify masked surgical instrument regions into the seven target categories after LoRA fine-tuning.
    Stage two premise; Table 1 reports the measured behavior, but the training label generation is left unspecified.
  • ad hoc to paper Instrument labels for SAM3-generated masks and 'No Tool' negatives can be obtained without pixel-level ground truth masks.
    The paper does not describe this process, yet the 'No Tool' category in Section 4.2 requires knowing which SAM3 masks are false positives. This is the load-bearing unstated assumption.
  • standard math The evaluation metrics Ch_IoU, ISI_IoU, and mc_IoU computed as in [10] are valid for the reported comparisons.
    Metrics are adopted from the cited ISINet work without re-derivation.
  • domain assumption The public dataset annotations and splits (EndoVis 2017, EndoVis 2018, GraSP) are used correctly.
    Reproduction relies on these external benchmarks and their labels.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Toward Mask Annotation-Free Surgical Instrument Segmentation from Endoscopic Images Using Text-Prompted Segment Anything Model 3 (SAM3)." pith.science (2026). https://pith.science/paper/3PSDDTOF

@misc{pith2026260808844,
  author       = {Pith},
  title        = {Pith review of: Toward Mask Annotation-Free Surgical Instrument Segmentation from Endoscopic Images Using Text-Prompted Segment Anything Model 3 (SAM3)},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/3PSDDTOF}},
  note         = {Machine review of arXiv:2608.08844}
}
read the original abstract

Surgical instrument segmentation is a fundamental task for computer-assisted interventions, yet most existing methods rely on pixel-level annotations or manual spatial prompts, which limit scalability and automation. The recently introduced Segment Anything Model 3 (SAM3) offers a pathway to annotation-free, automatic segmentation via text-based prompting; however, the instrument name as a text prompt could not be directly used due to a large domain gap. To overcome these limitations, we propose a two-stage framework that achieves instance-level segmentation without requiring ground truth masks or manual interaction. In the first stage, we leverage a natural-language-aligned generic prompt - "tool" - to produce binary masks using SAM3's zero-shot capability. In the second stage, these masks are extended to instance-level by integrating a vision-language model (Qwen) that is fine-tuned on SAM3-generated masked regions for instrument classification. We evaluate our approach on the EndoVis 2017 and 2018 datasets. Results show that, while our two-stage approach does not reach the performance of current fully supervised methods, it significantly outperforms the direct use of SAM3 for instance-level instrument segmentation with text prompts. Overall, our findings highlight both the limitations and potential of SAM3, suggesting a promising direction toward annotation-free surgical instrument segmentation.

Figures

Figures reproduced from arXiv: 2608.08844 by the authors.

Figure 1
Figure 1. Overview of the proposed pipeline. SAM3 generates binary masks from [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Illustration of segmentation outputs with and without mask merging. [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Plot of IoU vs. confidence threshold for the baseline model on En￾doVis 2018 dataset. We observe that using SAM3 baseline with the same confidence threshold t re￾sults in many missing predictions. There￾fore, we experimented with lowering the segmentation confidence threshold. The plot of IoU versus threshold is shown in [PITH_FULL_IMAGE:figures/full_fig_p009_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Instance-level segmentation of surgical instruments. The first column [PITH_FULL_IMAGE:figures/full_fig_p010_4.png]
Figure 5
Figure 5. Figure 5: Examples of failure cases in instance-level segmentation of surgical in [PITH_FULL_IMAGE:figures/full_fig_p011_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

30 extracted references · 16 canonical work pages

  1. [1]

    Artificial Intelligence Review58(1), 1 (2024)

    Ahmed, F.A., Yousef, M., Ahmed, M.A., Ali, H.O., Mahboob, A., Ali, H., Shah, Z., Aboumarzouk, O., Al Ansari, A., Balakrishnan, S.: Deep learning for surgical instrument recognition and segmentation in robotic-assisted surgeries: a systematic review. Artificial Intelligence Review58(1), 1 (2024)

    work page 2024

  2. [2]

    arXiv preprint arXiv:2001.11190 (2020)

    Allan,M.,Kondo,S.,Bodenstedt,S.,Leger,S.,Kadkhodamohammadi,R.,Luengo, I., Fuentes, F., Flouty, E., Mohammed, A., Pedersen, M., et al.: 2018 robotic scene segmentation challenge. arXiv preprint arXiv:2001.11190 (2020)

    arXiv 2020

  3. [3]

    arXiv preprint arXiv:1902.06426 (2019)

    Allan, M., Shvets, A., Kurmann, T., Zhang, Z., Duggal, R., Su, Y.H., Rieke, N., Laina, I., Kalavakonda, N., Bodenstedt, S., et al.: 2017 robotic instrument segmen- tation challenge. arXiv preprint arXiv:1902.06426 (2019)

    arXiv 2019

  4. [4]

    In: 2023 IEEE 20th Interna- tional Symposium on Biomedical Imaging (ISBI)

    Ayobi, N., Pérez-Rondón, A., Rodríguez, S., Arbeláez, P.: Matis: Masked-attention transformers for surgical instrument segmentation. In: 2023 IEEE 20th Interna- tional Symposium on Biomedical Imaging (ISBI). pp. 1–5. IEEE (2023)

    2023

  5. [5]

    Medical Image Analysis p

    Ayobi, N., Rodríguez, S., Pérez, A., Hernández, I., Aparicio, N., Dessevres, E., Peña, S., Santander, J., Caicedo, J.I., Fernández, N., et al.: Pixel-wise recognition for holistic surgical scene understanding. Medical Image Analysis p. 103726 (2025)

    work page 2025

  6. [6]

    In: Proceedings of the IEEE/CVF winter conference on applications of computer vision

    Baby, B., Thapar, D., Chasmai, M., Banerjee, T., Dargan, K., Suri, A., Banerjee, S., Arora, C.: From forks to forceps: A new framework for instance segmentation of surgical instruments. In: Proceedings of the IEEE/CVF winter conference on applications of computer vision. pp. 6191–6201 (2023)

    work page 2023

  7. [7]

    arXiv preprint arXiv:2502.13923 (2025)

    Bai, S., Chen, K., Liu, X., Wang, J., Ge, W., Song, S., Dang, K., Wang, P., Wang, S., Tang, J., et al.: Qwen2.5-VL technical report. arXiv preprint arXiv:2502.13923 (2025)

    arXiv 2025

  8. [8]

    arXiv preprint arXiv:2511.16719 (2025)

    Carion, N., Gustafson, L., Hu, Y.T., Debnath, S., Hu, R., Suris, D., Ryali, C., Alwala,K.V.,Khedr,H.,Huang,A.,etal.:Sam3:Segmentanythingwithconcepts. arXiv preprint arXiv:2511.16719 (2025)

    arXiv 2025

Show all 30 references
  1. [9]

    Medical Image Analysis81, 102569 (2022) 14 N

    Cerón, J.C.Á., Ruiz, G.O., Chang, L., Ali, S.: Real-time instance segmentation of surgical instruments using attention and multi-scale feature fusion. Medical Image Analysis81, 102569 (2022) 14 N. Poudelet al

    2022

  2. [10]

    González, C., Bravo-Sánchez, L., Arbelaez, P.: Isinet: an instance-based approach forsurgicalinstrumentsegmentation.In:Internationalconferenceonmedicalimage computing and computer-assisted intervention. pp. 595–605. Springer (2020)

    2020

  3. [11]

    In: Conference proceedings: Annual International Conference of the IEEE Engineering in Medicine and Biology Society

    Hasan,S.K.,Linte,C.A.:UNetPlus:Amodifiedencoder-decoderUNetarchitecture for semantic and instance segmentation of surgical instruments from laparoscopic images. In: Conference proceedings: Annual International Conference of the IEEE Engineering in Medicine and Biology Society....

    2019

  4. [12]

    ICLR1(2), 3 (2022)

    Hu, E.J., Shen, Y., Wallis, P., Allen-Zhu, Z., Li, Y., Wang, S., Wang, L., Chen, W., et al.: Lora: Low-rank adaptation of large language models. ICLR1(2), 3 (2022)

    2022

  5. [13]

    arXiv preprint arXiv:2510.08668 (2025)

    Jiang, S., Wang, Y., Song, S., Hu, T., Zhou, C., Pu, B., Zhang, Y., Yang, Z., Feng, Y., Zhou, J.T., et al.: Hulu-med: A transparent generalist model towards holistic medical vision-language understanding. arXiv preprint arXiv:2510.08668 (2025)

    2025

  6. [14]

    In: Interna- tional conference on medical image computing and computer-assisted intervention

    Jin, Y., Cheng, K., Dou, Q., Heng, P.A.: Incorporating temporal prior from motion flow for instrument segmentation in minimally invasive surgery video. In: Interna- tional conference on medical image computing and computer-assisted intervention. pp. 440–448. Springer (2019)

    2019

  7. [15]

    In: Proceedings of the IEEE/CVF international conference on computer vision

    Kirillov, A., Mintun, E., Ravi, N., Mao, H., Rolland, C., Gustafson, L., Xiao, T., Whitehead, S., Berg, A.C., Lo, W.Y., et al.: Segment Anything. In: Proceedings of the IEEE/CVF international conference on computer vision. pp. 4015–4026 (2023)

    2023

  8. [16]

    arXiv preprint arXiv:2502.06756 (2025)

    Lin, Y., Li, H., Shao, W., Yang, Z., Zhao, J., He, X., Luo, P., Zhang, K.: Samrefiner: Taming segment anything model for universal mask refinement. arXiv preprint arXiv:2502.06756 (2025)

    2025 arXiv

  9. [17]

    In: Medical Imaging 2025: Image Processing

    Lou, A., Li, Y., Zhang, Y., Labadie, R.F., Noble, J.: Zero-shot surgical tool segmen- tation in monocular video using Segment Anything Model 2. In: Medical Imaging 2025: Image Processing. vol. 13406, pp. 718–723. SPIE (2025)

    2025

  10. [18]

    In: Proceedings of the AAAI Conference on Artificial Intelligence

    Ni, Z.L., Bian, G.B., Wang, G.A., Zhou, X.H., Hou, Z.G., Chen, H.B., Xie, X.L.: Pyramid attention aggregation network for semantic segmentation of surgical in- struments. In: Proceedings of the AAAI Conference on Artificial Intelligence. vol. 34, pp. 11782–11790 (2020)

    2020

  11. [19]

    arXiv preprint arXiv:2601.16895 (2026)

    Poudel, N., Simon, R., Linte, C.A.: Evaluating large vision-language models for surgical tool detection. arXiv preprint arXiv:2601.16895 (2026)

    2026

  12. [20]

    arXiv preprint arXiv:2408.00714 (2024)

    Ravi, N., Gabeur, V., Hu, Y.T., Hu, R., Ryali, C., Ma, T., Khedr, H., Rädle, R., Rolland, C., Gustafson, L., et al.: (sam 2): Segment anything in images and videos. arXiv preprint arXiv:2408.00714 (2024)

    2024 arXiv

  13. [21]

    Rochester Institute of Technology: Research computing services (2026).https: //doi.org/10.34788/0S3G-QD15,https://www.rit.edu/researchcomputing/

    2026 doi

  14. [22]

    In: International Conference on Medical image computing and computer-assisted intervention

    Ronneberger, O., Fischer, P., Brox, T.: UNet: Convolutional networks for biomedi- cal image segmentation. In: International Conference on Medical image computing and computer-assisted intervention. pp. 234–241. Springer (2015)

    2015

  15. [23]

    International Journal of Computer Assisted Radiology and Surgery19(7), 1267–1271 (2024)

    Sheng, Y., Bano, S., Clarkson, M.J., Islam, M.: Surgical-DeSAM: decoupling sam for instrument segmentation in robotic surgery. International Journal of Computer Assisted Radiology and Surgery19(7), 1267–1271 (2024)

    2024

  16. [24]

    In: 2018 17th IEEE international conference on machine learning and applications (ICMLA)

    Shvets, A.A., Rakhlin, A., Kalinin, A.A., Iglovikov, V.I.: Automatic instrument segmentation in robot-assisted surgery using deep learning. In: 2018 17th IEEE international conference on machine learning and applications (ICMLA). pp. 624–

    2018

  17. [25]

    In: International conference on medical image computing and computer-assisted intervention

    Wang, A., Islam, M., Xu, M., Zhang, Y., Ren, H.: SAM meets robotic surgery: an empirical study on generalization, robustness and adaptation. In: International conference on medical image computing and computer-assisted intervention. pp. 234–244. Springer (2023) Mask Annotation...

    2023

  18. [26]

    In: Proceedings of the AAAI Conference on Artificial Intelligence

    Yue, W., Zhang, J., Hu, K., Xia, Y., Luo, J., Wang, Z.: Surgicalsam: Efficient class promptable surgical instrument segmentation. In: Proceedings of the AAAI Conference on Artificial Intelligence. vol. 38, pp. 6890–6898 (2024)

    2024

  19. [27]

    arXiv preprint arXiv:2304.13785 (2023)

    Zhang, K., Liu, D.: Customized segment anything model for medical image seg- mentation. arXiv preprint arXiv:2304.13785 (2023)

    2023 arXiv

  20. [28]

    Jour- nal of Imaging11(10), 364 (2025)

    Zhang, T., Yuan, X., Xu, H.: Surgical instrument segmentation via segment-then- classify framework with instance-level spatiotemporal consistency modeling. Jour- nal of Imaging11(10), 364 (2025)

    2025

  21. [29]

    In: 2022 International conference on robotics and automation (ICRA)

    Zhao, Z., Jin, Y., Heng, P.A.: Trasetr: track-to-segment transformer with con- trastive query for instance-level instrument segmentation in robotic surgery. In: 2022 International conference on robotics and automation (ICRA). pp. 11186– 11193. IEEE (2022)

    2022

  22. [30]

    Advances in Neural Infor- mation Processing Systems36, 28611–28623 (2023)

    Zhou, Z., Alabi, O., Wei, M., Vercauteren, T., Shi, M.: Text promptable surgical instrument segmentation with vision-language models. Advances in Neural Infor- mation Processing Systems36, 28611–28623 (2023)

    2023

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.