Pith. sign in

REVIEW 3 major objections 5 minor 30 references

Decoupling geometric deformation from semantic evolution in 4D Gaussian splatting produces continuous, text-promptable surgical scene understanding under tissue change.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

Decoupling geometry and semantics in 4DGS via HexPlane kinematic latents and rasterization-native extraction raises surgical semantic mIoU from 53.46% to 68.20% on CholecSeg8k.

T0 review reviewed 2026-07-11 challenge →

load-bearing objection Solid systems paper: dual-branch kinematic decoupling plus rasterization-native extraction delivers large, consistent mIoU gains on public surgical benchmarks; residual shared-encoder coupling is a real but secondary soft spot, not a collapse of the claim. the 3 major comments →

arxiv 2607.04761 v1 pith:VJISOI2C submitted 2026-07-06 cs.CV

DeGenseGS: Geometrically and Semantically Decoupled Surgical Scene Understanding in 4D Gaussian Splatting

Yimo Wang , Bin Kang , Shuojue Yang , Yueming Jin This is my paper
classification cs.CV
keywords 4D Gaussian SplattingSurgical Scene UnderstandingVision-Language ModelsSemantic-Geometric DecouplingHexPlaneText-promptable SegmentationRasterization-Native Extraction
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

The reading

Existing 4D Gaussian methods bind semantic features tightly to geometric warping, so when tissue appearance mutates in place (as in cauterization) the model invents false spatial displacements that destroy mask coherence. DeGenseGS instead extracts a shared kinematic latent from a HexPlane and routes it through two independent decoders—one for geometry, one for semantics—so appearance can update without forcing geometry to move. Semantics are then recovered from continuous 2D rasterized feature maps guided by the RGB image, and features are optimized by angular alignment on the hypersphere. On CholecSeg8k and EndoVis18 the method reaches 68.20 % and 44.63 % mean IoU, keeping spatially continuous labels through drastic deformation and topology change. A reader who wants autonomous robotic surgery to reason about anatomy in real time cares because reliable text-promptable 4D maps are a prerequisite for safety zones and instrument tracking.

Core claim

The paper establishes that independently modeling semantic evolution and geometric deformation—by routing a single HexPlane kinematic latent through separate geometry and semantic branches, then extracting labels from rasterized continuous feature maps—eliminates gradient-driven spurious warping and yields state-of-the-art semantic-anatomic alignment in dynamic endoscopic scenes.

What carries the argument

Kinematics-conditioned latent disentanglement: a multi-resolution HexPlane produces a shared latent Z_kin that is decoded by independent heads Ψ_geo and Ψ_sem, so semantic mutations can occur while geometric displacement stays near zero; this is paired with rasterization-native extraction on 2D feature maps.

Load-bearing premise

The dominant failure of coupled models is that semantic losses overwhelm geometric ones, and a single shared kinematic latent is sufficient to keep the two streams synchronized without introducing new inconsistencies.

What would settle it

On cauterization sequences, measure the actual 3D displacement of Gaussians before versus after appearance change; if DeGenseGS still produces large unfounded warps or if the mIoU gain disappears when the same architecture is run with frozen geometry, the central claim fails.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

If this is right

  • Text-promptable 4D reconstructions can maintain continuous anatomic labels through topology-altering surgical events.
  • Safety-critical zone alerting and instrument tracking become feasible on the reconstructed field rather than on 2D frames alone.
  • Gradient interference between appearance and geometry no longer forces a trade-off between visual fidelity and semantic coherence.
  • The same dual-branch routing can be applied whenever a scene exhibits in-place appearance mutations without physical motion.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same kinematic-latent split could stabilize multi-task 4DGS outside surgery whenever lighting, material, or staining changes dominate geometric motion.
  • If angular distillation on the hypersphere is the key regularizer, replacing the VLM backbone should still preserve most of the mIoU gain provided the feature space remains unit-norm.
  • Ablating the rasterization-native stage alone would quantify how much of the continuity comes from 2D graph filtering versus the 4D decoupling itself.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. DeGenseGS is a 4D Gaussian Splatting framework for text-promptable surgical scene understanding that aims to decouple geometric deformation from semantic evolution. A HexPlane encoder produces a shared kinematic latent Z_kin (Eqs. 1–2), which is decoded by independent branches Ψ_geo and Ψ_sem so that appearance mutations (e.g., cauterization) need not induce spurious spatial warping. Semantics are further refined by Angular-Aligned Semantic Distillation (cosine alignment on the hypersphere plus superpixel smoothness) and by Rasterization-Native Semantic Extraction (superpixel aggregation of rendered feature maps followed by RGB-guided filtering). On five CholecSeg8k and two EndoVis18 sequences the method reports 68.20% and 44.63% mIoU, respectively, outperforming SurgTPGS and several LangSplat/OpenGaussian variants; an ablation attributes most of the gain to the combined “S-G Decoupling” flag.

Significance. Real-time, queryable 4D reconstruction is a recognized bottleneck for autonomous robotic surgery. Explicitly separating semantic and geometric update pathways is a clean conceptual contribution relative to rigidly coupled Flexible Deformation Models, and the reported absolute gains (+14.74 mIoU on CholecSeg8k) are large enough to matter if they hold under broader evaluation. The rasterization-native extraction idea is practically useful for hole-free masks under topological change. The work therefore has clear clinical relevance provided the claimed isolation mechanism is shown to be the true source of the gains.

major comments (3)
  1. §2.2, Eqs. (1)–(2) and Fig. 1: The central claim is that independent decoding of a shared Z_kin eliminates gradient interference so that Δf_sem ≠ 0 while ΔG → 0 under in-place appearance change. Because both heads still back-propagate into the identical HexPlane encoder H(p,t) and base network, residual coupling is possible. The manuscript never measures residual geometric displacement (e.g., mean ||Δμ|| or ||ΔΣ|| on cauterization frames) under semantic-only loss, nor does it compare against an ablated encoder that freezes geometric parameters after the shared latent. Without such a diagnostic the isolation claim remains untested.
  2. Table 3: “S-G Decoupling” is a single binary that simultaneously enables both the dual-branch latent design (§2.2) and Rasterization-Native Semantic Extraction (§2.4). Consequently the +13 mIoU jump cannot be attributed to the kinematic disentanglement itself versus the 2-D post-process. A factorial ablation (dual-branch alone / RNSE alone / both) is required to support the paper’s strongest mechanistic claim.
  3. §3.1–3.3 and Tables 1–2: Results are reported on only seven sequences without multi-seed variance, confidence intervals, or statistical tests. Several per-class entries (e.g., gallbladder 12_15750, instrument-wrist Seq_5) show dramatic swings; without error bars it is unclear whether the average 68.20% / 44.63% figures are stable. At minimum, three random seeds and a paired significance test against SurgTPGS should be supplied.
minor comments (5)
  1. §2.3, Eq. (3): L_align is written as an expectation of 1 − ⟨·,·⟩; the text never states whether this is averaged over all pixels, only foreground pixels, or a sampled subset. Clarifying the sampling would aid reproducibility.
  2. §2.4: The unsupervised superpixel algorithm and the guided-filter radius are free parameters listed only in the implementation paragraph; their values (or a sensitivity plot) should appear in the main text or supplement.
  3. Fig. 3: Qualitative panels lack zoomed insets on the topology-altering regions (cauterization, instrument–tissue contact) that the introduction emphasizes; adding them would make the continuity claim more convincing.
  4. Related-work placement: FeatureEndo-4DGS and EndoWave are cited but their semantic-handling strategies are not contrasted with the proposed dual-branch design; a short paragraph would strengthen novelty positioning.
  5. Typographical: “spatiotemporal entanglement module” in the abstract versus “Kinematics-Conditioned Latent Disentanglement” in §2.2; consistent naming would reduce reader friction.

Circularity Check

0 steps flagged

No significant circularity: empirical 4DGS architecture evaluated on external public benchmarks with standard losses; no prediction reduces to its inputs by construction.

full rationale

DeGenseGS is an empirical computer-vision architecture paper. Its load-bearing claims are (i) that independent Ψ_geo / Ψ_sem branches conditioned on a shared HexPlane kinematic latent Z_kin (Eqs. 1–2) reduce spurious geometric warping relative to coupled Flexible Deformation Models, and (ii) that the resulting system yields higher mIoU on CholecSeg8k and EndoVis18. Neither claim is definitional: Z_kin, ΔG_t and Δf_sem,t are free network outputs optimized by ordinary reconstruction and angular-alignment losses (Eq. 3); the reported mIoU numbers are measured against independently published ground-truth masks on public sequences and compared to external baselines (LangSplat, OpenGaussian, SurgTPGS, etc.). No parameter is fitted to a quantity that is later re-presented as a prediction of that same quantity. Self-citations (e.g., Deform3DGS) appear only as related-work baselines that the paper argues against, not as uniqueness theorems that force the present design. Residual gradient flow through the shared HexPlane encoder is a methodological open question, not a circular reduction. Consequently the derivation chain contains no self-definitional step, no fitted-input-called-prediction, and no load-bearing self-citation chain.

Axiom & Free-Parameter Ledger

4 free parameters · 5 axioms · 3 invented entities

The central claim rests on standard 3DGS/HexPlane machinery, the domain premise that semantic and geometric gradients conflict under topology-preserving appearance change, and several new architectural modules whose necessity is shown only by ablation on the same datasets. No free physical constants; free parameters are ordinary deep-learning hyper-parameters and architectural choices.

free parameters (4)
  • Adam learning rate = 1.6e-3
    Set to 1.6e-3; controls optimization trajectory of both geometry and semantic branches.
  • deformation freeze iterations = 3000
    First 3 000 iterations freeze deformation parameters (coarse-to-fine schedule); chosen by hand.
  • kinematic latent dimension d_kin
    Dimension of Z_kin after concatenation of property-specific projections; not numerically specified but required for the dual-branch design.
  • superpixel count / guided-filter radius
    Hyper-parameters of the Region-aware Similarity Aggregation and Guided Filter in RNSE; affect boundary recovery but not listed.
axioms (5)
  • domain assumption 3D Gaussian Splatting alpha-blending and densification/pruning rules hold for dynamic endoscopic scenes.
    Preliminaries §2.1; inherited from Kerbl et al. and subsequent surgical 4DGS works.
  • domain assumption A multi-resolution HexPlane encoder can extract a shared spatiotemporal feature h_st that is informative for both geometry and semantics.
    §2.2; standard HexPlane assumption applied to the surgical setting.
  • domain assumption During topology-altering events such as cauterization, physical geometry remains largely static while appearance/semantics mutate, and coupled FDMs produce spurious warps because semantic gradients dominate.
    Core motivation in §1 and Fig. 1; treated as observed fact without quantitative gradient-norm evidence.
  • domain assumption VLM features live on a hypersphere so that angular (1 − cosine) distillation is the correct metric.
    §2.3 Angular-Aligned Semantic Distillation; common in open-vocabulary literature.
  • ad hoc to paper Independent decoding heads Ψ_geo and Ψ_sem from a shared Z_kin are sufficient to eliminate gradient interference without additional orthogonality constraints.
    Eqs. 1–2; architectural postulate validated only by ablation.
invented entities (3)
  • Kinematics-Conditioned Latent Disentanglement (shared Z_kin + dual Ψ_geo / Ψ_sem) no independent evidence
    purpose: Synchronize semantic mutations with scene dynamics while preventing semantic losses from warping geometry.
    Defined in §2.2; no independent evidence outside the paper’s own ablations.
  • Rasterization-Native Semantic Extraction (RNSE) no independent evidence
    purpose: Infer hole-free semantics from continuous 2D feature maps with RGB structural guidance rather than from 3D primitives.
    §2.4; combination of superpixel aggregation + guided filter tailored to the pipeline.
  • Angular-Aligned Semantic Distillation (AASD) no independent evidence
    purpose: Regularize VLM latent space via hyperspherical cosine loss plus superpixel smoothness.
    §2.3; minor variant of standard contrastive / cosine distillation.

reviewed 2026-07-11 · how reviews work

0 comments
Cite this review

Pith. "Pith review of DeGenseGS: Geometrically and Semantically Decoupled Surgical Scene Understanding in 4D Gaussian Splatting." pith.science (2026). https://pith.science/paper/VJISOI2C

@misc{pith2026260704761,
  author       = {Pith},
  title        = {Pith review of: DeGenseGS: Geometrically and Semantically Decoupled Surgical Scene Understanding in 4D Gaussian Splatting},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/VJISOI2C}},
  note         = {Machine review of arXiv:2607.04761}
}
Share X Bluesky LinkedIn Reddit HN
read the original abstract

Real-time, text-promptable 4D reconstruction is indispensable for autonomous surgical interaction. Severe misalignment between semantic meaning and physical anatomy still persists, largely because existing solutions integrate Vision-Language Models into deformable fields via a rigid coupling scheme that tightly binds semantic features to geometric warping. In this paper, we propose DeGenseGS, Geometrically and Semantically Decoupled Surgical Scene Understanding in 4D Gaussian Splatting, a novel framework that independently models semantic evolution and geometric deformation. Specifically, we propose a HexPlane-based spatiotemporal entanglement module that uses shared kinematic latents to synchronize semantic mutations with scene dynamics, while explicitly disentangling semantic updates from geometric deformation. To further ensure robustness against reconstruction artifacts, we devise a Rasterization-Native Semantic Extraction mechanism that infers semantics from topologically continuous feature maps. Additionally, we incorporate an angular-aligned optimization strategy that conforms to the native hyperspherical latent space, thereby preventing semantic distortion. Extensive evaluations on the CholecSeg8k and EndoVis18 datasets demonstrate that DeGenseGS achieves state-of-the-art performance. Our framework yields enhanced geometric completeness and robust semantic-anatomic alignment, enabling spatially continuous segmentation despite drastic tissue deformation and topological transitions.

Figures

Figures reproduced from arXiv: 2607.04761 by Bin Kang, Shuojue Yang, Yimo Wang, Yueming Jin.

Figure 1
Figure 1. Figure 1: Illustration of spatiotemporal feature misalignment in existing coupled 4D Gaussian Splatting frameworks. (A) Physical Reality: During topology-altering events such as cauterization, tissue appearance mutates significantly while the un￾derlying geometry remains largely stable. (B) Limitation of existing Coupled Frameworks: Due to the gradient sensitivity variation, the coupled framework leads to unfounded … view at source ↗
Figure 2
Figure 2. Figure 2: Overview of the proposed DeGenseGS framework. To prevent spurious geomet￾ric warping, a HexPlane encoder extracts shared kinematic latents to enable indepen￾dent geometric and semantic decoding , thereby explicitly decoupling the optimiza￾tion pathways. Subsequently, Angular-Aligned Semantic Distillation, together with a Rasterization-Native Semantic Extraction mechanism, strengthens feature grounding and … view at source ↗
Figure 3
Figure 3. Figure 3: Qualitative result on CholecSeg8k and EndoVis18 datasets [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

30 extracted references · 4 linked inside Pith

  1. [1]

    arXiv preprint arXiv:2001.11190 (2020)

    Allan, M., Kondo, S., Bodenstedt, S., Leger, S., Kadkhodamohammadi, R., Lu- engo, I., et al.: 2018 Robotic Scene Segmentation Challenge. arXiv preprint arXiv:2001.11190 (2020)

    Pith/arXiv arXiv 2018

  2. [2]

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

    Chen, J., Zhang, X., Hoque, M.I., Vasconcelos, F., Stoyanov, D., Elson, D.S., Huang, B.: SurgicalGS: Dynamic 3D Gaussian Splatting for Accurate Robotic- Assisted Surgical Scene Reconstruction. In: International Conference on Medical Image Computing and Computer-Assisted Intervention. pp. 572–582. Springer Na- ture Switzerland (2025)

    2025

  3. [3]

    In: IEEE/CVF International Conference on Computer Vision

    Chen, Q., Yang, L., Chen, Y., Zhao, N., Lai, J., Shao, J., Xie, X.: Training-Free Class Purification for Open-Vocabulary Semantic Segmentation. In: IEEE/CVF International Conference on Computer Vision. pp. 23124–23134. IEEE (2025)

    2025

  4. [4]

    In: IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Cho, S., Shin, H., Hong, S., Arnab, A., Seo, P.H., Kim, S.: CAT-Seg: Cost Aggre- gation for Open-Vocabulary Semantic Segmentation. In: IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 4113–4123 (2024)

    2024

  5. [5]

    arXiv preprint arXiv:2012.12453 (2020)

    Hong, W.Y., Kao, C.L., Kuo, Y.H., Wang, J.R., Chang, W.L., Shih, C.S.: Cholec- Seg8k: A Semantic Segmentation Dataset for Laparoscopic Cholecystectomy Based on Cholec80. arXiv preprint arXiv:2012.12453 (2020)

    Pith/arXiv arXiv 2012

  6. [6]

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

    Huang, Y., Bai, L., Cui, B., Yuan, K., Wang, G., Hoque, M.I., et al.: SurgTPGS: Semantic 3D Surgical Scene Understanding with Text Promptable Gaussian Splat- ting. In: International Conference on Medical Image Computing and Computer- Assisted Intervention. pp. 584–594. Springer (2025) 10 Y. Wang et al

    2025

  7. [7]

    Kerbl, B., Kopanas, G., Leimkühler, T., Drettakis, G.: 3D Gaussian Splatting for Real-TimeRadianceFieldRendering.ACMTransactionsonGraphics42(4),139–1 (2023)

    2023

  8. [8]

    In: IEEE/CVF International Conference on Computer Vision

    Kerr, J., Kim, C.M., Goldberg, K., Kanazawa, A., Tancik, M.: LERF: Language Embedded Radiance Fields. In: IEEE/CVF International Conference on Computer Vision. pp. 19729–19739. IEEE (2023)

    2023

  9. [9]

    arXiv preprint arXiv:2502.14886 (2025)

    Khan, U., Nawaz, U., Qayyum, A., Ashraf, S., Xie, Y., Khan, M.H., Qadir, J.: Sur- gical Scene Understanding in the Era of Foundation AI Models: A Comprehensive Review. arXiv preprint arXiv:2502.14886 (2025)

    arXiv 2025

  10. [10]

    In: IEEE/CVF International Conference on Computer Vision

    Kirillov, A., Mintun, E., Ravi, N., Mao, H., Rolland, C., Gustafson, L., et al.: Segment Anything. In: IEEE/CVF International Conference on Computer Vision. pp. 4015–4026. IEEE (2023)

    2023

  11. [11]

    In: European Conference on Computer Vision

    Labe, I., Issachar, N., Lang, I., Benaim, S.: DGD: Dynamic 3D Gaussians Distilla- tion. In: European Conference on Computer Vision. pp. 361–378. Springer (2024)

    2024

  12. [12]

    arXiv preprint arXiv:2408.07981 (2024)

    Li,J.,Skinner,G.,Yang,G.,Quaranto,B.R.,Schwaitzberg,S.D.,Kim,P.C.,Xiong, J.: LLaVA-Surg: Towards Multimodal Surgical Assistant via Structured Surgical Video Learning. arXiv preprint arXiv:2408.07981 (2024)

    Pith/arXiv arXiv 2024

  13. [13]

    In: Machine Learning for Health Symposium

    Li, K., Wang, J., Han, W., Zhao, D.: FeatureEndo-4DGS: Real-Time Deformable Surgical Scene Reconstruction and Segmentation with 4D Gaussian Splatting. In: Machine Learning for Health Symposium. pp. 1218–1234. PMLR (2026)

    2026

  14. [14]

    In: IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Li, Z., Chen, Z., Li, Z., Xu, Y.: Spacetime Gaussian Feature Splatting for Real- Time Dynamic View Synthesis. In: IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 8508–8520. IEEE (2024)

    2024

  15. [15]

    IEEE Transactions on Medical Imaging43(4), 1308–1322 (2023)

    Liu, M., Han, Y., Wang, J., Wang, C., Wang, Y., Meijering, E.: LSKANet: Long Strip Kernel Attention Network for Robotic Surgical Scene Segmentation. IEEE Transactions on Medical Imaging43(4), 1308–1322 (2023)

    2023

  16. [16]

    arXiv preprint arXiv:2401.12561 (2024)

    Liu, Y., Li, C., Yang, C., Yuan, Y.: EndoGaussian: Real-Time Gaussian Splatting for Dynamic Endoscopic Scene Reconstruction. arXiv preprint arXiv:2401.12561 (2024)

    Pith/arXiv arXiv 2024

  17. [17]

    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

  18. [18]

    In: IEEE/CVF Conference on Computer Vision and Pattern Recogni- tion

    Qin, M., Li, W., Zhou, J., Wang, H., Pfister, H.: LangSplat: 3D Language Gaussian Splatting. In: IEEE/CVF Conference on Computer Vision and Pattern Recogni- tion. pp. 20051–20060. IEEE (2024)

    2024

  19. [19]

    Shan,J.,Cai,Z.,Hsieh,C.T.,Han,L.,Cheng,S.S.,Wang,H.:DeformableGaussian SplattingforEfficientandHigh-FidelityReconstructionofSurgicalScenes.In:2025 IEEE International Conference on Robotics and Automation (ICRA). pp. 10545– 10551. IEEE (2025)

    2025

  20. [20]

    In: ICLR 2025 Workshop on Foundation Models in the Wild (2025)

    Wang, G., Bai, L., Nah, W.J., Wang, J., Zhang, Z., Chen, Z., et al.: Surgical-LVLM: Learning to Adapt Large Vision-Language Model for Grounded Visual Question Answering in Robotic Surgery. In: ICLR 2025 Workshop on Foundation Models in the Wild (2025)

    2025

  21. [21]

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

    Wang, Y., Long, Y., Fan, S.H., Dou, Q.: Neural Rendering for Stereo 3D Recon- struction of Deformable Tissues in Robotic Surgery. In: International Conference on Medical Image Computing and Computer-Assisted Intervention. pp. 431–441. Springer (2022)

    2022

  22. [22]

    arXiv preprint arXiv:2510.23087 (2025) Title Suppressed Due to Excessive Length 11

    Wu, T., Miao, Y., Guo, J., Chen, Z., Zhao, S., Li, Z., et al.: EndoWave: Rational- Wavelet 4D Gaussian Splatting for Endoscopic Reconstruction. arXiv preprint arXiv:2510.23087 (2025) Title Suppressed Due to Excessive Length 11

    arXiv 2025

  23. [23]

    In: Advances in Neural Information Processing Systems

    Wu, Y., Meng, J., Li, H., Wu, C., Shi, Y., Cheng, X., et al.: OpenGaussian: Towards Point-Level 3D Gaussian-Based Open Vocabulary Understanding. In: Advances in Neural Information Processing Systems. vol. 37, pp. 19114–19138 (2024)

    2024

  24. [24]

    In: International Con- ference on Medical Image Computing and Computer-Assisted Intervention

    Yang, C., Wang, K., Wang, Y., Yang, X., Shen, W.: Neural LerPlane Represen- tations for Fast 4D Reconstruction of Deformable Tissues. In: International Con- ference on Medical Image Computing and Computer-Assisted Intervention. pp. 46–56. Springer (2023)

    2023

  25. [25]

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

    Yang, S., Li, Q., Shen, D., Gong, B., Dou, Q., Jin, Y.: Deform3DGS: Flexible Deformation for Fast Surgical Scene Reconstruction with Gaussian Splatting. In: International Conference on Medical Image Computing and Computer-Assisted Intervention. pp. 132–142. Springer (2024)

    2024

  26. [26]

    In: Advances in Neural Information Processing Systems

    Yu, T., Kumar, S., Gupta, A., Levine, S., Hausman, K., Finn, C.: Gradient Surgery for Multi-Task Learning. In: Advances in Neural Information Processing Systems. vol. 33, pp. 5824–5836 (2020)

    2020

  27. [27]

    Medical Image Analysis105, 103644 (2025)

    Yuan, K., Srivastav, V., Yu, T., Lavanchy, J.L., Marescaux, J., Mascagni, P., Padoy, N.: Learning Multi-Modal Representations by Watching Hundreds of Sur- gical Video Lectures. Medical Image Analysis105, 103644 (2025)

    2025

  28. [28]

    In: AAAI Conference on Artificial Intelligence

    Yue,W.,Zhang,J.,Hu,K.,Xia,Y.,Luo,J.,Wang,Z.:SurgicalSAM:EfficientClass Promptable Surgical Instrument Segmentation. In: AAAI Conference on Artificial Intelligence. vol. 38, pp. 6890–6898. AAAI (2024)

    2024

  29. [29]

    Computers in Biology and Medicine171, 108238 (2024)

    Zhang, Y., Shen, Z., Jiao, R.: Segment Anything Model for Medical Image Seg- mentation: Current Applications and Future Directions. Computers in Biology and Medicine171, 108238 (2024)

    2024

  30. [30]

    Pattern Recognition161, 111271 (2025)

    Zhao, C., Huang, X., Yang, K., Wang, X., Wang, Q.: Generalizable 3D Gaussian Splatting for Novel View Synthesis. Pattern Recognition161, 111271 (2025)

    2025

This paper was first reviewed by grok-4.5 on July 11, 2026.