What question did this study set out to answer?

Evaluate the efficiency of a 3D Gaussian Splatting segmentation pipeline for cultural heritage documentation.

Automatic Segmentation of 3D Gaussian Splatting in Cultural Heritage Complex

View Full Paper

Ask AI

Bookmark

Evaluate the efficiency of a 3D Gaussian Splatting segmentation pipeline for cultural heritage documentation.
Proposed an automated semantic segmentation pipeline integrating the Segment Anything Model 3 (SAM 3) with per-class prompting for Gaussian reconstruction.
Applied to a nadiral UAV dataset of the Siti Inggil heritage complex.
Assessed segmentation performance against manually segmented 2D and 3D reference data, supplemented by M3C2 geometric accuracy assessment.
Segmentation performance revealed promise across four classes: ground, roofs, vegetation, and water bodies.
Challenges observed related to acquisition geometry, surface characteristics, and foundational model limitations.
Geometric accuracy assessments highlighted discrepancies compared to manual segmentation.

No takes yet. Share an insight, caveat, or question.

Randomized trial evaluates automated semantic segmentation in cultural heritage, indicating potential and challenges.

Fadilah et al. (2026) studied this question.

View Full Paper

Ask AI

Bookmark