More than Segmentation: Benchmarking SAM 3 for Segmentation, 3D Perception, and Reconstruction in Robotic Surgery

TL;DR

This paper evaluates SAM 3 and SAM 3D in robotic surgery, demonstrating their improved segmentation and 3D perception capabilities, while highlighting current limitations and potential for future domain-specific enhancements.

Contribution

The paper provides a comprehensive benchmark of SAM 3 and SAM 3D in surgical scenarios, emphasizing their performance in zero-shot segmentation, language prompts, and 3D reconstruction.

Findings

SAM 3 outperforms previous models in image and video segmentation.

SAM 3D shows strong depth estimation and 3D reconstruction abilities.

Language prompts currently underperform in surgical contexts.

Abstract

The recent SAM 3 and SAM 3D have introduced significant advancements over the predecessor, SAM 2, particularly with the integration of language-based segmentation and enhanced 3D perception capabilities. SAM 3 supports zero-shot segmentation across a wide range of prompts, including point, bounding box, and language-based prompts, allowing for more flexible and intuitive interactions with the model. In this empirical evaluation, we assess the performance of SAM 3 in robot-assisted surgery, benchmarking its zero-shot segmentation with point and bounding box prompts and exploring its effectiveness in dynamic video tracking, alongside its newly introduced language prompt segmentation. While language prompts show potential, their performance in the surgical domain is currently suboptimal, highlighting the need for further domain-specific training. Additionally, we investigate SAM 3D's depth reconstruction abilities, demonstrating its capacity to process surgical scene data and reconstruct 3D anatomical structures from 2D images. Through comprehensive testing on the MICCAI EndoVis 2017 and EndoVis 2018 benchmarks, SAM 3 shows clear improvements over SAM and SAM 2 in both image and video segmentation under spatial prompts, while the zero-shot evaluations of SAM 3D on SCARED, StereoMIS, and EndoNeRF indicate strong monocular depth estimation and realistic 3D instrument reconstruction, yet also reveal remaining limitations in complex, highly dynamic surgical scenes.