ToothForge: Automatic Dental Shape Generation using Synchronized Spectral Embeddings

Tibor Kubík, François Guibault, Michal Španěl, Hervé Lombaert · 2025

Read ToothForge: Automatic Dental Shape Generation using Synchronized Spectral Embeddings by Tibor Kubík, François Guibault, Michal Španěl, Hervé Lombaert (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2506-02702.txt.

Description

Tibor Kubík and co-authors present ToothForge in 2025, a method that generates new three-dimensional teeth automatically. Dental shape data sets are small, which starves learning models. Their method works in the frequency domain, where a tooth shape becomes a compact spectrum that a model can learn and sample in milliseconds.

Raw spectra are unstable, since each shape's harmonics can twist differently. The team aligns all spectra to a shared basis before training, which removes that bias and frees the method from requiring identical mesh wiring. On real dental crowns, aligned training rebuilt shapes more faithfully than unaligned. The authors also show shape compression and blending, and note uses beyond dentistry.

Catalog record

CREATORTibor Kubík, François Guibault, Michal Španěl, Hervé Lombaert
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE31 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2506-02702

Browse the full library · About the Tower of Babel