Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

A 3D Facial Reconstruction Evaluation Methodology: Comparing Smartphone Scans with Deep Learning Based Methods Using Geometry and Morphometry Criteria

Álvaro Heredia-Lidón, Alejandro Moñux-Bernal, Alejandro González, Luis M. Echeverry-Quiceno, Max Rubert, Aroa Casado, María Esther Esteban, Mireia Andreu-Montoriol, Susanna Gallardo, Cristina Ruffo, Neus Martínez-Abadías, Xavier Sevillano

TL;DR

The paper addresses the challenge of validating low-cost 3D facial acquisition and reconstruction methods by introducing a geometry-plus-morphometry framework. It combines traditional geometric metrics with geometric morphometrics ($GPA$ and $EDMA$) to quantify global and local shape preservation, using stereophotogrammetry ($SPG$) ground-truth models. In a case study, smartphone-based scans and deep-learning reconstructions from 2D images are evaluated against SPG, revealing that smartphone scans generally better preserve facial morphology than the DL reconstructions, though static capture limits practicality. The proposed methodology provides a robust, biology-informed validation tool that can guide improvements in low-cost facial imaging technologies for clinical applications.

Abstract

Three-dimensional (3D) facial shape analysis has gained interest due to its potential clinical applications. However, the high cost of advanced 3D facial acquisition systems limits their widespread use, driving the development of low-cost acquisition and reconstruction methods. This study introduces a novel evaluation methodology that goes beyond traditional geometry-based benchmarks by integrating morphometric shape analysis techniques, providing a statistical framework for assessing facial morphology preservation. As a case study, we compare smartphone-based 3D scans with state-of-the-art deep learning reconstruction methods from 2D images, using high-end stereophotogrammetry models as ground truth. This methodology enables a quantitative assessment of global and local shape differences, offering a biologically meaningful validation approach for low-cost 3D facial acquisition and reconstruction techniques.

A 3D Facial Reconstruction Evaluation Methodology: Comparing Smartphone Scans with Deep Learning Based Methods Using Geometry and Morphometry Criteria

TL;DR

The paper addresses the challenge of validating low-cost 3D facial acquisition and reconstruction methods by introducing a geometry-plus-morphometry framework. It combines traditional geometric metrics with geometric morphometrics ( and ) to quantify global and local shape preservation, using stereophotogrammetry () ground-truth models. In a case study, smartphone-based scans and deep-learning reconstructions from 2D images are evaluated against SPG, revealing that smartphone scans generally better preserve facial morphology than the DL reconstructions, though static capture limits practicality. The proposed methodology provides a robust, biology-informed validation tool that can guide improvements in low-cost facial imaging technologies for clinical applications.

Abstract

Three-dimensional (3D) facial shape analysis has gained interest due to its potential clinical applications. However, the high cost of advanced 3D facial acquisition systems limits their widespread use, driving the development of low-cost acquisition and reconstruction methods. This study introduces a novel evaluation methodology that goes beyond traditional geometry-based benchmarks by integrating morphometric shape analysis techniques, providing a statistical framework for assessing facial morphology preservation. As a case study, we compare smartphone-based 3D scans with state-of-the-art deep learning reconstruction methods from 2D images, using high-end stereophotogrammetry models as ground truth. This methodology enables a quantitative assessment of global and local shape differences, offering a biologically meaningful validation approach for low-cost 3D facial acquisition and reconstruction techniques.

Paper Structure

This paper contains 16 sections, 7 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Materials and Methods
  3. Participants Recruitment and Ethical Statement
  4. Face Acquisition and Reconstruction
  5. High-end facial acquisition.
  6. Low-cost acquisition/reconstruction.
  7. Automatic landmarking
  8. Evaluation metrics
  9. Geometric evaluation metrics.
  10. Morphometric evaluation metrics.
  11. Results and discussion
  12. Experimental setup.
  13. Geometric evaluation
  14. Morphometric evaluation
  15. Conclusions
  16. ...and 1 more sections

Figures (7)

  • Figure 1: Methodology pipeline. (a) Data acquisition using high-cost 3D stereophotogrammetry system, and low-cost approaches. (b) Facial mesh extraction and normalisation. (c) 3D anatomical landmarks registration for morphometric analysis. (d) Methods comparison based on geometry and morphometry metrics.
  • Figure 2: Face acquisition and reconstruction. (a) High quality facial models acquired by a stereophotogrammetry (SPG) system: setup and reconstruction of the 3D facial model from 10 images. (b) Low-cost facial model acquisition via iPhoneX TrueDepth camera and processed 3D model. (c) Low-cost facial reconstruction from 2D image(s): 3DDFA_V3 Wang2024, HRN3 Lei2023 (3 images) and Era3D Li2024.
  • Figure 3: At-a-glance comparison of the geometry of the 3D facial acquisition and reconstruction methods.
  • Figure 4: Surface-to-surface desviation between low-cost methods and SPG. Deviation values are coded with a colour map from blue to red.
  • Figure 5: (a-d) Centroid Size (CS) and (e-h) Pairwise Procrustes Distances (PPD) between SPG and low-cost methods. The $r$ coefficient shows the correlation between variables.
  • ...and 2 more figures