IoSR: End-to-End Intraoral Scans Repairing


Manel Farhat (Digital Research Center of Sfax Laboratory of Signals, Systems, Artificial Intelligence and Networks Technopark of Sfax), Achraf Ben-hamadou (Digital Research Center of Sfax Laboratory of Signals, Systems, Artificial Intelligence and Networks Technopark of Sfax), Ahmed Rekik (Digital Research Center of Sfax Laboratory of Signals, Systems, Artificial Intelligence and Networks Technopark of Sfax), Ons Abida (Digital Research Center of Sfax Laboratory of Signals, Systems, Artificial Intelligence and Networks Technopark of Sfax), Oussama Smaoui (Biotech Dental Group)
The 35th British Machine Vision Conference
PDFPosterVideo (Right click to download)

Abstract

Despite being crucial, Intraoral 3D Scan Repairing (IoSR) is often overlooked in research, as it remains a labor-intensive, behind-the-scenes task in digital dental workflows. Scanning limitations stemming from scanning technology, intraoral anatomy, and device handling can lead to gaps or artifacts, making high-precision reconstruction essential for clinical procedures as aligners, surgical guides, and crowns. Efficient scan repair can also reduce the back-and-forth between dentists and dental labs, or even the need to reschedule additional scanning sessions, thus streamlining the treatment process. This paper presents the first comprehensive approach to address this challenge, introducing an end-to-end method for repairing intraoral 3D scans by leveraging both geometric and implicit representations. Our model employs a two-block architecture: a coarse generator with a geometric-aware transformer encoder and multilayer perceptron, followed a fine generator allows generating dense points and refining the generated point cloud based on implicit representation. Experimental results show that our method outperforms state-of-the-art approaches offering enhanced accuracy and generating point clouds with improved fidelity

Citation

@inproceedings{Farhat_2025_BMVC,
author    = {Manel Farhat and Achraf Ben-hamadou and Ahmed Rekik and Ons Abida and Oussama Smaoui},
title     = {IoSR: End-to-End Intraoral Scans Repairing},
booktitle = {36th British Machine Vision Conference 2025, {BMVC} 2025, Sheffield, UK, November 24-27, 2025},
publisher = {BMVA},
year      = {2025},
url       = {https://bmva-archive.org.uk/bmvc/2025/assets/papers/Paper_1196/paper.pdf}
}


Copyright © 2025 The British Machine Vision Association and Society for Pattern Recognition
The British Machine Vision Conference is organised by The British Machine Vision Association and Society for Pattern Recognition. The Association is a Company limited by guarantee, No.2543446, and a non-profit-making body, registered in England and Wales as Charity No.1002307 (Registered Office: Dept. of Computer Science, Durham University, South Road, Durham, DH1 3LE, UK).

Imprint | Data Protection