Diff2DGS: Reliable Reconstruction of Occluded Surgical Scenes via 2D Gaussian Splatting

• URL: http://arxiv.org/abs/2602.18314v1
• Date: Fri, 20 Feb 2026 16:14:21 GMT
• Title: Diff2DGS: Reliable Reconstruction of Occluded Surgical Scenes via 2D Gaussian Splatting
• Authors: Tianyi Song, Danail Stoyanov, Evangelos Mazomenos, Francisco Vasconcelos,
• Abstract summary: We propose Diff2DGS, a novel two-stage framework for reliable 3D reconstruction of occluded surgical scenes.<n>In the first stage, a diffusion-based video module with temporal priors inpaints tissue occluded by instruments with high spatial-temporal consistency.<n>In the second stage, we adapt 2D Gaussian Splatting (2DGS) with a Learnable Deformation Model (LDM) to capture dynamic tissue deformation and anatomical geometry.<n>D Diff2DGS outperforms state-of-the-art approaches in both appearance and geometry, reaching 38.02 dB PSNR on
• Score: 10.70948053935438
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Real-time reconstruction of deformable surgical scenes is vital for advancing robotic surgery, improving surgeon guidance, and enabling automation. Recent methods achieve dense reconstructions from da Vinci robotic surgery videos, with Gaussian Splatting (GS) offering real-time performance via graphics acceleration. However, reconstruction quality in occluded regions remains limited, and depth accuracy has not been fully assessed, as benchmarks like EndoNeRF and StereoMIS lack 3D ground truth. We propose Diff2DGS, a novel two-stage framework for reliable 3D reconstruction of occluded surgical scenes. In the first stage, a diffusion-based video module with temporal priors inpaints tissue occluded by instruments with high spatial-temporal consistency. In the second stage, we adapt 2D Gaussian Splatting (2DGS) with a Learnable Deformation Model (LDM) to capture dynamic tissue deformation and anatomical geometry. We also extend evaluation beyond prior image-quality metrics by performing quantitative depth accuracy analysis on the SCARED dataset. Diff2DGS outperforms state-of-the-art approaches in both appearance and geometry, reaching 38.02 dB PSNR on EndoNeRF and 34.40 dB on StereoMIS. Furthermore, our experiments demonstrate that optimizing for image quality alone does not necessarily translate into optimal 3D reconstruction accuracy. To address this, we further optimize the depth quality of the reconstructed 3D results, ensuring more faithful geometry in addition to high-fidelity appearance.

Related papers

• ERGO: Excess-Risk-Guided Optimization for High-Fidelity Monocular 3D Gaussian Splatting [63.138778159026934]
  We propose an adaptive optimization framework guided by excess risk decomposition, termed ERGO.<n> ERGO dynamically estimates the view-specific excess risk and adaptively adjust loss weights during optimization.<n>Experiments on the Google Scanned Objects dataset and the OmniObject3D dataset demonstrate the superiority of ERGO over existing state-of-the-art methods.
  arXiv  Detail & Related papers  (2026-02-10T20:44:43Z)
• SAGS: Self-Adaptive Alias-Free Gaussian Splatting for Dynamic Surgical Endoscopic Reconstruction [10.831257597641669]
  We propose SAGS, a self-adaptive alias-free Gaussian splatting framework for deformable tissue reconstruction.<n>We introduce an attention-driven, dynamically weighted 4D deformation decoder, leveraging 3D smoothing filters and 2D Mip filters to mitigate artifacts in deformable tissue reconstruction.
  arXiv  Detail & Related papers  (2025-10-31T09:42:32Z)
• 2DGS-R: Revisiting the Normal Consistency Regularization in 2D Gaussian Splatting [21.933933778984954]
  We present 2DGS-R, a new method that uses a hierarchical training approach to improve rendering quality while maintaining geometric accuracy.<n>Compared to the original 2DGS, our method requires only 1% more storage and minimal additional training time.<n>These findings indicate that our approach effectively balances efficiency with performance, leading to improvements in both visual fidelity and geometric reconstruction accuracy.
  arXiv  Detail & Related papers  (2025-10-19T13:52:29Z)
• GauSSmart: Enhanced 3D Reconstruction through 2D Foundation Models and Geometric Filtering [50.675710727721786]
  We propose GauSSmart, a hybrid method that bridges 2D foundational models and 3D Gaussian Splatting reconstruction.<n>Our approach integrates established 2D computer vision techniques, including convex filtering and semantic feature supervision.<n>We validate our approach across three datasets, where GauSSmart consistently outperforms existing Gaussian Splatting.
  arXiv  Detail & Related papers  (2025-10-16T03:38:26Z)
• 3D Gaussian Flats: Hybrid 2D/3D Photometric Scene Reconstruction [62.84879632157956]
  We propose a novel hybrid 2D/3D representation that jointly optimize constrained planar (2D) Gaussians for modeling flat surfaces and freeform (3D) Gaussians for the rest of the scene.<n>Our end-to-end approach dynamically detects and refines planar regions, improving both visual fidelity and geometric accuracy.<n>It achieves state-of-the-art depth estimation on ScanNet++ and ScanNetv2, and excels at mesh extraction without overfitting to a specific camera model.
  arXiv  Detail & Related papers  (2025-09-19T21:04:36Z)
• EndoLRMGS: Complete Endoscopic Scene Reconstruction combining Large Reconstruction Modelling and Gaussian Splatting [16.50682401904587]
  We propose EndoLRMGS, that combines Large Reconstruction Modelling (LRM) and Gaussian Splatting (GS) for complete surgical scene reconstruction.<n>GS reconstructs deformable tissues and LRM generates 3D models for surgical tools while position and scale are subsequently optimized.<n>In experiments on four surgical videos from three public datasets, our method improves the Intersection-over-union (IoU) of tool 3D models in 2D projections by>40%.
  arXiv  Detail & Related papers  (2025-03-28T13:57:12Z)
• T-3DGS: Removing Transient Objects for 3D Scene Reconstruction [83.05271859398779]
  Transient objects in video sequences can significantly degrade the quality of 3D scene reconstructions.<n>We propose T-3DGS, a novel framework that robustly filters out transient distractors during 3D reconstruction using Gaussian Splatting.
  arXiv  Detail & Related papers  (2024-11-29T07:45:24Z)
• SurgicalGS: Dynamic 3D Gaussian Splatting for Accurate Robotic-Assisted Surgical Scene Reconstruction [18.074890506856114]
  We present SurgicalGS, a dynamic 3D Gaussian Splatting framework specifically designed for surgical scene reconstruction with improved geometric accuracy. Our approach first initialises a Gaussian point cloud using depth priors, employing binary motion masks to identify pixels with significant depth variations and fusing point clouds from depth maps across frames for initialisation. We use the Flexible Deformation Model to represent dynamic scene and introduce a normalised depth regularisation loss along with an unsupervised depth smoothness constraint to ensure more accurate geometric reconstruction.
  arXiv  Detail & Related papers  (2024-10-11T22:46:46Z)
• A Review of 3D Reconstruction Techniques for Deformable Tissues in Robotic Surgery [8.909938295090827]
  NeRF-based techniques have recently garnered attention for the ability to reconstruct scenes implicitly. On the other hand, 3D-GS represents scenes explicitly using 3D Gaussians and projects them onto a 2D plane as a replacement for the complex volume rendering in NeRF. This work explores and reviews state-of-the-art (SOTA) approaches, discussing their innovations and implementation principles.
  arXiv  Detail & Related papers  (2024-08-08T12:51:23Z)
• 3D Vessel Reconstruction from Sparse-View Dynamic DSA Images via Vessel Probability Guided Attenuation Learning [79.60829508459753]
  Current commercial Digital Subtraction Angiography (DSA) systems typically demand hundreds of scanning views to perform reconstruction. The dynamic blood flow and insufficient input of sparse-view DSA images present significant challenges to the 3D vessel reconstruction task. We propose to use a time-agnostic vessel probability field to solve this problem effectively.
  arXiv  Detail & Related papers  (2024-05-17T11:23:33Z)
• 2D Gaussian Splatting for Geometrically Accurate Radiance Fields [50.056790168812114]
  3D Gaussian Splatting (3DGS) has recently revolutionized radiance field reconstruction, achieving high quality novel view synthesis and fast rendering speed without baking.<n>We present 2D Gaussian Splatting (2DGS), a novel approach to model and reconstruct geometrically accurate radiance fields from multi-view images.<n>We demonstrate that our differentiable terms allows for noise-free and detailed geometry reconstruction while maintaining competitive appearance quality, fast training speed, and real-time rendering.
  arXiv  Detail & Related papers  (2024-03-26T17:21:24Z)
• Endo-4DGS: Endoscopic Monocular Scene Reconstruction with 4D Gaussian Splatting [12.333523732756163]
  Dynamic scene reconstruction can significantly enhance downstream tasks and improve surgical outcomes. NeRF-based methods have recently risen to prominence for their exceptional ability to reconstruct scenes. We present Endo-4DGS, a real-time endoscopic dynamic reconstruction approach.
  arXiv  Detail & Related papers  (2024-01-29T18:55:29Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.