SurfDist: Interpretable Three-Dimensional Instance Segmentation Using Curved Surface Patches

2025/07/11 by Borchardt, Jackson, Kato, Saul
#Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences

paper · doi:10.48550/arxiv.2507.08223

Abstract

We present SurfDist, a convolutional neural network architecture for three-dimensional volumetric instance segmentation. SurfDist enables prediction of instances represented as closed surfaces composed of smooth parametric surface patches, specifically bicubic Bézier triangles. SurfDist is a modification of the popular model architecture StarDist-3D which breaks StarDist-3D's coupling of instance parameterization dimension and instance voxel resolution, and it produces predictions which may be upsampled to arbitrarily high resolutions without introduction of voxelization artifacts. For datasets with blob-shaped instances, common in biomedical imaging, SurfDist can outperform StarDist-3D with more compact instance parameterizations. We detail SurfDist's technical implementation and show one synthetic and one real-world dataset for which it outperforms StarDist-3D. These results demonstrate that interpretable instance surface models can be learned effectively alongside instance membership.

Citations

• End-to-End Object Detection with Transformers
• Cellpose: a generalist algorithm for cellular segmentation
• SciPy 1.0--Fundamental Algorithms for Scientific Computing in Python
• 3D U-Net: Learning Dense Volumetric Segmentation from Sparse Annotation
• U-Net: Convolutional Networks for Biomedical Image Segmentation

Related

• Voxel2Mesh: 3D Mesh Model Generation from Volumetric Data
• AnyStar: Domain randomized universal star-convex 3D instance segmentation
• Vox-Surf: Voxel-based Implicit Surface Representation
• HIVE: HIerarchical Volume Encoding for Neural Implicit Surface Reconstruction
• MASC: Multi-scale Affinity with Sparse Convolution for 3D Instance Segmentation