Related Experiment Video
Updated: Jun 17, 2026
10:14
3D Scanning Technology Bridging Microcircuits and Macroscale Brain Images in 3D Novel Embedding Overlapping Protocol
Published on: May 12, 2019
3D face recognition using simulated annealing and the surface interpenetration measure.
Chauã C Queirolo1, Luciano Silva, Olga R P Bellon
1Departamento de Informatica, Universidade Federal do Parana, Caixa Postal 19092, Curitiba, PR 81531-980, Brazil. chaua@inf.ufpr.br
IEEE Transactions on Pattern Analysis and Machine Intelligence
|January 16, 2010
Summary
This study introduces an automatic 3D face recognition framework using Simulated Annealing (SA) and Surface Interpenetration Measure (SIM). The novel method achieves high accuracy in verification and identification, outperforming existing state-of-the-art results on the FRGC v2 database.
Area of Science:
- Computer Vision
- Biometrics
- Pattern Recognition
Background:
- 3D face recognition is crucial for security and identification systems.
- Existing methods struggle with variations in facial expressions and pose.
- Accurate and robust 3D face recognition frameworks are in high demand.
Purpose of the Study:
- To present a novel automatic framework for 3D face recognition.
- To improve the accuracy and robustness of 3D face recognition systems, especially under varying facial expressions.
- To achieve state-of-the-art performance on a large-scale 3D face database.
Main Methods:
- A Simulated Annealing (SA)-based approach for range image registration using Surface Interpenetration Measure (SIM) as a similarity metric.
- Combining SIM values from four distinct facial regions (nose, forehead, entire face) for authentication score.
- A modified SA approach incorporating invariant face regions to enhance handling of facial expressions.
Main Results:
- Achieved a 96.5% verification rate at a 0.1% False Acceptance Rate (FAR) on the FRGC v2 database.
- Attained a rank-one accuracy of 98.4% in the identification scenario.
- Demonstrated superior performance compared to existing state-of-the-art methods on the FRGC v2 database.
Conclusions:
- The proposed automatic 3D face recognition framework is highly effective and robust.
- The method demonstrates significant improvements in accuracy and handling of facial expressions.
- This work sets a new benchmark for 3D face recognition performance on the FRGC v2 database.
