TCATSEG: A TOOTH CENTER-WISE ATTENTION NETWORK FOR 3D DENTAL MODEL SEMANTIC SEGMENTATION
Qiang He, University of Chinese Academy of Sciences, Jiajia Dai, Changsong Lei, University of Chinese Academy of Sciences, Shaofeng Wang, Feifei Zuo, Yaqian Liang
Abstract
Accurate semantic segmentation of 3D dental models is essential for digital dentistry applications such as orthodontics and dental implants. However, due to complex tooth arrangements and similarities in shape among adjacent teeth, existing methods struggle with accurate segmentation, because they often focus on local geometry while neglecting global contextual information. To address this, we propose TCATSeg, a novel framework that combines local geometric features with global semantic context. We introduce a set of sparse yet physically meaningful superpoints to capture global semantic relationships and enhance segmentation accuracy. Additionally, we present a new dataset of 400 dental models, including pre-orthodontic samples, to evaluate the generalization of our method. Extensive experiments demonstrate that TCATSeg outperforms state-of-the-art approaches.
BibTeX
@inproceedings{icassp2026_tcatsegatoothcen,
title = {TCATSEG: A TOOTH CENTER-WISE ATTENTION NETWORK FOR 3D DENTAL MODEL SEMANTIC SEGMENTATION},
author = {Qiang He and University of Chinese Academy of Sciences and Jiajia Dai and Changsong Lei and University of Chinese Academy of Sciences and Shaofeng Wang and Feifei Zuo and Yaqian Liang and Xiaoming Deng and University of Chinese Academy of Sciences and Yong-Jin Liu and Hongan Wang and University of Chinese Academy of Sciences},
booktitle = {ICASSP 2026},
year = {2026}
}