Memory-Augmented SAM2 for Training-Free Surgical Video Segmentation

Fu Wang; Ming Yin; Xujiong Ye; Yanda Meng; Zeyu Fu

REVIEW 3 major objections 3 minor 29 references

Memory-Augmented SAM2 for Training-Free Surgical Video Segmentation

T0 review · 3 major / 3 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read A training-free memory split into context and occlusion branches lifts SAM2's surgical segmentation by 6.1% and 4.36% on two EndoVis benchmarks.

desk verdict A training-free SAM2 memory redesign that improves surgical instrument tracking on two small benchmarks; the gains are plausible but the hand-set thresholds and lack of error bars leave the size of the effect unproven. read the letter →

arxiv 2507.09577 v2 pith:ZDHRF5GY submitted 2025-07-13 cs.CV

Ming Yin , Fu Wang , Xujiong Ye , Yanda Meng , Zeyu Fu This is my paper

classification cs.CV

keywords surgicalinstrumentsegmentationvideoobjectSAM2training-freememory-augmentedtrackingocclusionhandlingcontext-awarememoryEndoVisbenchmarks

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

The paper claims that SAM2's greedy, first-in-first-out memory bank is the main reason it loses track of surgical instruments in long videos, and that a training-free memory redesign addresses this. It introduces MA-SAM2, which separates the memory into a context-aware branch that keeps high-quality historical masks and an occlusion-resilient branch that stores frames with strong interference. With no additional parameters and no training, MA-SAM2 improves Challenge IoU by 6.1 percent on EndoVis2017 and 4.36 percent on EndoVis2018 relative to SAM2. The practical point is that a fix to memory management, rather than a new model, can make foundation-model segmentation usable in the operating room.

What carries the argument

The load-bearing object is the augmented memory bank, split into two curated stores. The context-aware memory (CAM) carries the best-scoring historical masks selected by a cumulative log-IoU hypothesis test, which combats slow drift; the occlusion-resilient memory (ORM) carries up to five frames flagged as having strong interference, selected by an IoU-confidence threshold and a bounding-box overlap window, which resists sudden occlusion. The bank fills ORM first and CAM only when spare capacity remains, and the outputs of both branches are integrated to build the final segmentation representation.

What would settle it

Sweep the paper's hand-set values on EndoVis2017—IoU threshold from 0.5 to 0.95, bounding-box overlap window, and ORM capacity from 1 to 10—then fix the best values and run them on EndoVis2018 without retuning; if the reported 6.1% and 4.36% gains shrink or reverse, the thresholds were likely tuned to the test videos.

Watch

Extended reading notes

Core claim

The paper's central claim is that SAM2's sequential memory policy pollutes its tracking when intermediate frames are occluded or ambiguous, so the paper replaces that policy with two complementary memories. The context-aware memory scores three candidate masks per frame by a cumulative log-IoU and stores the best historical masks, while the occlusion-resilient memory screens each frame with an IoU-confidence threshold (at least 0.8) and a bounding-box overlap window (0.6–0.9) to identify interference frames, keeping up to five such frames. The two memories are stored in a fixed-capacity bank that prioritizes the occlusion-resilient memory and adds context-aware memory only when capacity remains. In multi-instrument videos, a single initial mask per category is propagated through a one-prompt, multi-target loop. The authors show on EndoVis2017 and EndoVis2018 that this memory split outperforms SAM2 and each memory branch individually, with the occlusion-resilient memory contributing the larger gain.

Load-bearing premise

The improvements hinge on several hand-picked values—a minimum overlap score (IoU) of 0.8, a bounding-box overlap between 0.6 and 0.9, a storage cap of 5 frames, and other unstated thresholds—whose choice is never justified or tested for sensitivity; if those values were tuned on the test videos, the reported gains could be artifacts.

Editorial extensions

If this is right

On both EndoVis2017 and EndoVis2018, the full MA-SAM2 beats plain SAM2, and the ablation attributes most of the gain to the occlusion-resilient memory rather than the context-aware memory.
The one-prompt, multi-target loop means a video sequence needs only one initial mask per instrument category; no corrective prompts are required during tracking.
Because the method is training-free and adds no parameters, it can be dropped into any existing SAM2-based segmentation pipeline as a memory-policy replacement.
The single-prompt design also speeds up inference relative to per-target processing, which matters for real-time surgical use.
The improvement is concentrated in some instrument categories such as Bipolar Forceps, Monopolar Curved Scissors, and Ultrasound Probe, indicating the memory policy helps most where occlusion and motion are frequent.

Reading between the lines

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

If the hand-set thresholds (IoU at least 0.8, bounding-box overlap in 0.6–0.9, ORM capacity 5, and the unnamed CAM thresholds and capacity) were tuned on the test videos, the benchmark gains could be selection artifacts; the paper does not report how the values were chosen or how sensitive the results are to them.
The same context-plus-occlusion memory split could transfer to other long-video object-tracking domains with frequent occlusion, such as sports or surveillance, since the mechanism does not depend on surgical appearance.
The large gain from ORM suggests that SAM2's practical failures in long videos stem more from its memory update rule than from its image features, which would encourage research on memory-scheduling policies for foundation trackers.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Desk Editor's Note

read the letter

Worth a read if you work on SAM2 memory or surgical tracking. The paper does what it says: it takes SAM2's memory bank, splits candidates into a context-aware branch (cumulative log-IoU scoring, after SAM2Long) and an occlusion-resilient branch (distractor-aware memory, after Videnovic et al., extended to multiple targets), and runs one-prompt multi-target inference. No extra parameters, no training. Reported gains over SAM2 are about 6 points on EndoVis2017 and 4.4 points on EndoVis2018, and the ablation shows most of that comes from the occlusion-resilient memory. The combination is genuinely new, and the one-prompt multi-target loop is a sensible practical contribution.

Soft spots, in order of weight. First, all evaluation is on 8 and 15 videos, with no error bars, no per-video variance, and no significance tests. On datasets this small, a tuned choice on the test split can move Challenge IoU by more than the reported gains. Second, the method's thresholds and capacities are hand-set and not justified: IoU confidence >= 0.8, bbox overlap 0.6-0.9, ORM capacity 5, plus unnamed CAM thresholds and capacity. No sensitivity analysis, no grid search, no held-out validation. The stress-test concern about selection artifacts is legitimate, especially because all videos are used for testing. Third, the abstract has the datasets reversed: it says 4.36% on EndoVis2017 and 6.1% on EndoVis2018, while the tables and Section 1 show the opposite. Fourth, Eq. (2) doesn't match its prose: the text says alternative masks are processed by removing overlap with the primary mask, but the formula keeps the overlap and just applies connected components to the non-overlapping part of Ms, so the result is almost Ms itself. That looks like a typo, but it needs fixing.

The core claim, that a memory redesign can improve SAM2 without retraining, is supported by the tables and the ablation. The magnitude and generalizability are not fully established because of the tuning and small-test issues, but this is an addressable problem, not a load-bearing flaw. The paper is honest about its scope, the comparison to SAM2Long and distractor-aware memory is fair, and the code is promised.

For a referee: yes, I would send it out. It is a clean, testable, incremental contribution that belongs in a surgical segmentation venue. The review should ask for error bars, a sensitivity analysis on the thresholds, a correction of the abstract and Eq. (2), and ideally a transfer check to another long-video dataset. If those come back clean, this is a usable result.

Referee Report

3 major / 3 minor

Summary. The paper proposes MA-SAM2, a training-free modification of SAM2 for surgical video segmentation. It replaces SAM2's greedy FIFO memory with two complementary memory modules: a context-aware memory (CAM) that stores historical high-quality masks through cumulative-score hypothesis selection, and an occlusion-resilient memory (ORM) that filters contaminated masks using IoU and bounding-box overlap criteria. The method is evaluated on EndoVis2017 and EndoVis2018 under zero-shot settings, reporting Challenge IoU improvements of 6.10 and 4.36 points over SAM2, with an ablation showing that ORM provides most of the gain. The code is publicly released.

Significance. If the reported results hold, MA-SAM2 offers a genuinely simple, training-free improvement over SAM2 on two public surgical benchmarks, which is practically useful and reproducible because no learned parameters are introduced and the code is available. The ablation study is a strength: it clearly attributes the main gain to the ORM module rather than to the baseline alone. The significance is, however, moderated by the small test sets and by the fact that several hand-set hyperparameters are neither derived nor tested for sensitivity, so the consistency and generalizability of the gains are not yet established.

major comments (3)

[§2.2, Tables 1–2] The reported gains of 6.10 and 4.36 Challenge IoU points rest on a set of hand-set hyperparameters that are neither derived nor validated: the IoU confidence threshold θ_IoU ≥ 0.8, the bounding-box overlap range 0.6–0.9, the ORM maximum capacity of 5 frames, the unnamed CAM confidence and IoU thresholds, the unnamed CAM maximum capacity, and the constant ε in Eq. (3). Because all 8 EndoVis2017 and 15 EndoVis2018 videos are used for testing and no sensitivity analysis is reported, the improvements in Tables 1 and 2 could result from selecting these thresholds on the exact test split. The paper should report how each threshold was chosen, provide a sweep or sensitivity study around each value, and ideally include a transfer check with thresholds fixed on one dataset and evaluated on the other. The claim of being 'without additional parameters' should also be qualified: no parameters are learned, but several manual hyperparameters are introduced.
[§3, Tables 1–3] No measure of variability or statistical significance is provided for any number in Tables 1–3. With only 8 and 15 test videos, the gap between MA-SAM2 and SAM2 could be within video-level variance, and the ablation increments for CAM (0.89 points on EndoVis2018 and 1.47 points on EndoVis2017) are especially small relative to likely video-level noise. The manuscript should report per-video Challenge IoU scores, error bars or bootstrap confidence intervals, and a paired significance test (e.g., Wilcoxon signed-rank or a paired bootstrap) for both the full-model comparison and the ablation rows.
[§2.2, Eqs. (2)–(3)] Equation (2) is ambiguous and appears internally inconsistent with the prose: the text says alternative masks are post-processed by removing regions that overlap with Ms and retaining only their largest connected components, but the displayed formula includes the full overlap region Ms ∩ Ma and applies the CC operator only to Ms \ (Ms ∩ Ma). Since this formula defines the core occlusion-resilient filter, which the ablation shows is the main source of the improvement, it must be unambiguous. Similarly, Eq. (3) writes S(t − 1) on the right-hand side although the left-hand side is S_k(t); the cumulative score should be per-branch, i.e., S_k(t − 1), otherwise the CAM hypothesis-selection rule is not reproducible. Please correct the notation and define the special cases where the overlap region or the non-overlap region is empty.

minor comments (3)

[Abstract] The abstract states performance improvements of 4.36% and 6.1% on EndoVis2017 and EndoVis2018, respectively, but Section 1 and Tables 1–2 report 6.10% on EndoVis2017 and 4.36% on EndoVis2018; the order in the abstract should be reversed.
[§1 and §2.1] There are several phrasing errors, for example 'a mask-based one-prompt strategy provides a single prompt' reads as a sentence fragment, and 'The SAM2 process each input frame' should be 'SAM2 processes each input frame'.
[§1 and §3] The paper emphasizes real-time requirements but reports no runtime or latency comparison against SAM2 or Surgical SAM2; consider adding inference-time measurements or softening the real-time claim.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity found; the memory-augmentation heuristics are evaluated on external public benchmarks and are not derived from the reported metrics.

full rationale

The paper's central claim is an empirical performance comparison: MA-SAM2 improves Challenge IoU over SAM2 on EndoVis2017 and EndoVis2018 without training. This claim is not a first-principles derivation from the method's equations; it is a benchmark result against external baselines. The memory design is a heuristic extension of prior work by independent groups (Videnovic et al. and Ding et al.), with no load-bearing self-citation and no uniqueness theorem imported from the authors' own prior publications. The update rules, including Eq. (2) for the interference mask and Eq. (3) for cumulative candidate scores, are selection heuristics over decoder outputs rather than definitions that reintroduce the evaluation metric as an input. No equation in the paper defines the reported improvement in terms of the memory strategy's own parameters, and no fitted parameter is renamed as a prediction. The hand-set thresholds (theta_IoU >= 0.8, bounding-box overlap 0.6-0.9, ORM capacity 5) are unvalidated hyperparameters that could affect generalization if tuned on the test videos, but the paper does not report such tuning, and the absence of sensitivity analysis is a robustness concern rather than a circular derivation. Therefore no specific circular step can be quoted and exhibited as required.

Assumptions & free parameters 6 free parameters · 5 assumptions · 0 invented entities

Free parameters are hand-set thresholds and capacities in the memory design; their values are partly reported in the text, but CAM thresholds and capacities are left unspecified. Axioms are the background assumptions needed for the benchmark claim: SAM2's pretrained model, correct first-frame prompts, the suitability of the EndoVis datasets, and the validity of applying a single-sequence scoring rule to multi-target masks. No new physical entities are introduced; CAM and ORM are organizational modules rather than independently evidenced new objects.

free parameters (6)

IoU confidence threshold theta_IoU = 0.8
Used to reject uncertain masks before storing them in occlusion-resilient memory; no derivation or sensitivity analysis is reported (Section 2.2).
Bounding-box overlap range = 0.6 to 0.9
Used to decide whether a frame contains significant interference; no validation or sensitivity analysis is reported (Section 2.2).
ORM maximum capacity = 5 frames
Storage limit set to maintain inference speed; no ablation across different capacities is reported (Section 2.2).
CAM confidence and IoU thresholds = not specified
Preset thresholds for storing frames in context-aware memory; the actual values are not given in the paper (Section 2.2).
CAM maximum capacity = not specified
Dynamic maximum capacity for context-aware memory is mentioned but never quantified (Section 2.2).
epsilon in cumulative score = small constant
Added in Eq. (3) to avoid taking the logarithm of zero; the value is not reported (Section 2.2).

assumptions (5)

domain assumption SAM2 provides a usable video segmentation backbone with pretrained weights.
MA-SAM2 builds on SAM2's memory-attention architecture and pretrained model without modifying parameters (Section 2).
domain assumption The first-frame mask prompt for each instrument category is correct and complete.
The one-prompt strategy initializes tracking from the first-frame masks, so errors in those prompts would propagate through the sequence (Section 2.2).
domain assumption SAM2's greedy FIFO memory is the cause of tracking failures in surgical videos.
This motivates the CAM and ORM design but is supported only by qualitative examples in Fig. 1, not by a controlled diagnosis (Section 1).
domain assumption EndoVis2017 and EndoVis2018 instrument-type labels and evaluation protocol are appropriate for the claimed comparison.
The paper reuses public datasets with labels from Gonzalez et al. and does not discuss label noise, annotation ambiguity, or protocol limitations (Section 3).
domain assumption Cumulative log-IoU scoring from SAM2Long remains valid when extended to multi-target masks.
The paper adopts Ding et al.'s Eq. (3) without deriving why the same scoring logic applies when each candidate mask contains multiple instrument instances (Section 2.2).

how reviews work

0 comments

Cite this review

Pith. "Pith review of Memory-Augmented SAM2 for Training-Free Surgical Video Segmentation." pith.science (2026). https://pith.science/paper/ZDHRF5GY

@misc{pith2026250709577,
  author       = {Pith},
  title        = {Pith review of: Memory-Augmented SAM2 for Training-Free Surgical Video Segmentation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ZDHRF5GY}},
  note         = {Machine review of arXiv:2507.09577}
}

read the original abstract

Surgical video segmentation is a critical task in computer-assisted surgery, essential for enhancing surgical quality and patient outcomes. Recently, the Segment Anything Model 2 (SAM2) framework has demonstrated remarkable advancements in both image and video segmentation. However, the inherent limitations of SAM2's greedy selection memory design are amplified by the unique properties of surgical videos-rapid instrument movement, frequent occlusion, and complex instrument-tissue interaction-resulting in diminished performance in the segmentation of complex, long videos. To address these challenges, we introduce Memory Augmented (MA)-SAM2, a training-free video object segmentation strategy, featuring novel context-aware and occlusion-resilient memory models. MA-SAM2 exhibits strong robustness against occlusions and interactions arising from complex instrument movements while maintaining accuracy in segmenting objects throughout videos. Employing a multi-target, single-loop, one-prompt inference further enhances the efficiency of the tracking process in multi-instrument videos. Without introducing any additional parameters or requiring further training, MA-SAM2 achieved performance improvements of 4.36% and 6.1% over SAM2 on the EndoVis2017 and EndoVis2018 datasets, respectively, demonstrating its potential for practical surgical applications.

Figures

Figures reproduced from arXiv: 2507.09577 by the authors.

**Figure 1.** Performance comparison of occlusion handling and temporal consistency between MA-SAM2 and SAM2: (a) Tracking over time during instrument reappearance, (b) Tracking over time during multiple instrument overlap. prompts, SAM generates precise segmentation masks for objects within static images. This remarkable zero-shot learning capability has been validated by numerous studies [18,11], as exemplified by PerSAM [28]… view at source ↗

**Figure 2.** Architecture of the proposed model MA-SAM2. memory (CAM) and occlusion-resilient memory (ORM). Specifically, CAM utilises a collaborative hypothesis pruning mechanism to maintain optimal target clues during long-term video inference, thereby mitigating the impact of erroneous predictions and resolving the issue of lost reappearance identification. ORM employs a variation selection mechanism to identify and suppress… view at source ↗

**Figure 3.** Memory bank comparison between the SAM2 and our MA-SAM2. ing the spatial dimensions of the image embedding. The goal is to generate a segmentation mask M(c, t) for each instrument category c ∈ C at frame t. The segmentation is performed by a model e.g., SAM2, formulated as: M(c, t) = SAM2({I1, ..., IN }, c, t) (1) As shown in [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗

Figures from the paper (1 more)

**Figure 4.** Figure 4: Qualitative Comparison on the EndoVis2018 Dataset [PITH_FULL_IMAGE:figures/full_fig_p008_4.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

29 extracted references · 13 canonical work pages

[1]

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
[2]

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
[3]

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)

work page 2023
[4]

arXiv preprint arXiv:2410.16268 (2024)

Ding, S., Qian, R., Dong, X., Zhang, P., Zang, Y., Cao, Y., Guo, Y., Lin, D., Wang, J.: Sam2long: Enhancing sam 2 for long video segmentation with a training-free memory tree. arXiv preprint arXiv:2410.16268 (2024)

arXiv 2024
[5]

In: International Conference on Medical Im- age Computing and Computer-Assisted Intervention

González, C., Bravo-Sánchez, L., Arbelaez, P.: Isinet: an instance-based approach for surgical instrument segmentation. In: International Conference on Medical Im- age Computing and Computer-Assisted Intervention. pp. 595–605. Springer (2020)

work page 2020
[6]

In: 2020 Digital Image Computing: Techniques and Applications (DICTA)

Jian, Z., Yue, W., Wu, Q., Li, W., Wang, Z., Lam, V.: Multitask learning for video- based surgical skill assessment. In: 2020 Digital Image Computing: Techniques and Applications (DICTA). pp. 1–8. IEEE (2020)

work page 2020
[7]

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]

arXiv preprint arXiv:2408.07931 (2024)

Liu, H., Zhang, E., Wu, J., Hong, M., Jin, Y.: Surgical sam 2: Real-time segment anything in surgical video by efficient frame pruning. arXiv preprint arXiv:2408.07931 (2024)

arXiv 2024

Show all 29 references

[9]

Nature Communications15(1), 654 (2024)

Ma, J., He, Y., Li, F., Han, L., You, C., Wang, B.: Segment anything in medical images. Nature Communications15(1), 654 (2024)

2024
[10]

Medical Image Analysis76, 102310 (2022)

Ni, Z.L., Zhou, X.H., Wang, G.A., Yue, W.Q., Li, Z., Bian, G.B., Hou, Z.G.: Surginet: Pyramid attention aggregation and class-wise self-distillation for surgical instrument segmentation. Medical Image Analysis76, 102310 (2022)

2022
[11]

Knowledge-Based Systems p

Peng, Y., Lin, X., Ma, N., Du, J., Liu, C., Liu, C., Chen, Q.: Sam-lad: Segment any- thing model meets zero-shot logic anomaly detection. Knowledge-Based Systems p. 113176 (2025)

2025
[12]

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.: Sam2: Segment anything in images and videos. arXiv preprint arXiv:2408.00714 (2024)

2024 arXiv
[13]

Science translational medicine 8(337), 337ra64–337ra64 (2016)

Shademan, A., Decker, R.S., Opfermann, J.D., Leonard, S., Krieger, A., Kim, P.C.: Supervised autonomous robotic soft tissue surgery. Science translational medicine 8(337), 337ra64–337ra64 (2016)

2016
[14]

arXiv preprint arXiv:2408.04098 (2024)

Shen, Y., Ding, H., Shao, X., Unberath, M.: Performance and non-adversarial robustness of the segment anything model 2 in surgical video segmentation. arXiv preprint arXiv:2408.04098 (2024)

2024 arXiv
[15]

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
[16]

Biomedical Signal Processing and Control 102, 107296 (2025) 10 M.Yin et al

Song, M., Zhai, C., Yang, L., Liu, Y., Bian, G.: An attention-guided multi-scale fusion network for surgical instrument segmentation. Biomedical Signal Processing and Control 102, 107296 (2025) 10 M.Yin et al

2025
[17]

arXiv preprint arXiv:2411.17576 (2024)

Videnovic, J., Lukezic, A., Kristan, M.: A distractor-aware memory for visual ob- ject tracking with sam2. arXiv preprint arXiv:2411.17576 (2024)

2024 arXiv
[18]

Smart Agricultural Technology 8, 100515 (2024)

Williams, D., Macfarlane, F., Britten, A.: Leaf only sam: A segment anything pipeline for zero-shot automated leaf segmentation. Smart Agricultural Technology 8, 100515 (2024)

2024
[19]

arXiv preprint arXiv:2304.12620 (2023)

Wu, J., Ji, W., Liu, Y., Fu, H., Xu, M., Xu, Y., Jin, Y.: Medical sam adapter: Adapting segment anything model for medical image segmentation. arXiv preprint arXiv:2304.12620 (2023)

2023 arXiv
[20]

arXiv preprint arXiv:2408.03286 (2024)

Yan, Z., Sun, W., Zhou, R., Yuan, Z., Zhang, K., Li, Y., Liu, T., Li, Q., Li, X., He, L., et al.: Biomedical sam 2: Segment anything in biomedical images and videos. arXiv preprint arXiv:2408.03286 (2024)

2024 arXiv
[21]

arXiv preprint arXiv:2304.11968 (2023)

Yang, J., Gao, M., Li, Z., Gao, S., Wang, F., Zheng, F.: Track anything: Segment anything meets videos. arXiv preprint arXiv:2304.11968 (2023)

2023 arXiv
[22]

Computers in Biology and Medicine 151, 106216 (2022)

Yang,L.,Gu,Y.,Bian,G.,Liu,Y.:Anattention-guidednetworkforsurgicalinstru- ment segmentation from endoscopic images. Computers in Biology and Medicine 151, 106216 (2022)

2022
[23]

IEEE Transactions on Medical Robotics and Bionics5(2), 323–334 (2023)

Yang, L., Wang, H., Gu, Y., Bian, G., Liu, Y., Yu, H.: Tma-net: A transformer- based multi-scale attention network for surgical instrument segmentation. IEEE Transactions on Medical Robotics and Bionics5(2), 323–334 (2023)

2023
[24]

arXiv preprint arXiv:2408.04593 (2024)

Yu, J., Wang, A., Dong, W., Xu, M., Islam, M., Wang, J., Bai, L., Ren, H.: Sam 2 in robotic surgery: An empirical evaluation for robustness and generalization in surgical video segmentation. arXiv preprint arXiv:2408.04593 (2024)

2024 arXiv
[25]

IEEE transactions on medical imag- ing 42(10), 2817–2831 (2023)

Yue, W., Liao, H., Xia, Y., Lam, V., Luo, J., Wang, Z.: Cascade multi-level trans- former network for surgical workflow analysis. IEEE transactions on medical imag- ing 42(10), 2817–2831 (2023)

2023
[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
[27]

IEEE Internet of Things Journal 8(10), 7789–7817 (2020)

Zhang, J., Tao, D.: Empowering things with intelligence: a survey of the progress, challenges, and opportunities in artificial intelligence of things. IEEE Internet of Things Journal 8(10), 7789–7817 (2020)

2020
[28]

arXiv preprint arXiv:2305.03048 (2023)

Zhang, R., Jiang, Z., Guo, Z., Yan, S., Pan, J., Ma, X., Dong, H., Gao, P., Li, H.: Personalize segment anything model with one shot. arXiv preprint arXiv:2305.03048 (2023)

2023 arXiv
[29]

arXiv preprint arXiv:2408.00874 (2024)

Zhu, J., Qi, Y., Wu, J.: Medical sam 2: Segment medical images as video via segment anything model 2. arXiv preprint arXiv:2408.00874 (2024)

2024 arXiv

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.

[1] [1]

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

[2] [2]

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

[3] [3]

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)

work page 2023

[4] [4]

arXiv preprint arXiv:2410.16268 (2024)

Ding, S., Qian, R., Dong, X., Zhang, P., Zang, Y., Cao, Y., Guo, Y., Lin, D., Wang, J.: Sam2long: Enhancing sam 2 for long video segmentation with a training-free memory tree. arXiv preprint arXiv:2410.16268 (2024)

arXiv 2024

[5] [5]

In: International Conference on Medical Im- age Computing and Computer-Assisted Intervention

González, C., Bravo-Sánchez, L., Arbelaez, P.: Isinet: an instance-based approach for surgical instrument segmentation. In: International Conference on Medical Im- age Computing and Computer-Assisted Intervention. pp. 595–605. Springer (2020)

work page 2020

[6] [6]

In: 2020 Digital Image Computing: Techniques and Applications (DICTA)

Jian, Z., Yue, W., Wu, Q., Li, W., Wang, Z., Lam, V.: Multitask learning for video- based surgical skill assessment. In: 2020 Digital Image Computing: Techniques and Applications (DICTA). pp. 1–8. IEEE (2020)

work page 2020

[7] [7]

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] [8]

arXiv preprint arXiv:2408.07931 (2024)

Liu, H., Zhang, E., Wu, J., Hong, M., Jin, Y.: Surgical sam 2: Real-time segment anything in surgical video by efficient frame pruning. arXiv preprint arXiv:2408.07931 (2024)

arXiv 2024