Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

3D Face Morphing Attack Generation using Non-Rigid Registration

Jag Mohan Singh, Raghavendra Ramachandra

TL;DR

The paper tackles morphing attacks on face recognition by introducing the first direct 3D morphing method that operates on 3D face point clouds. It registers two bona fide faces using Bayesian Coherent Point Drift (BCPD), then averages geometry and color to produce a morph, forming a three-step pipeline: 3D-3D alignment, colorization, and 3D morphing. Evaluated on the Facescape dataset with 200 identities, the method achieves high attack potential, with GMAP values up to 97.93% for 3D FRS and 100% for 2D FRS, surpassing the prior state-of-the-art and highlighting the need for robust 3D morphing detection. The work demonstrates the feasibility and strength of 3D morphing attacks and underscores important directions for improving FRS security against such threats.

Abstract

Face Recognition Systems (FRS) are widely used in commercial environments, such as e-commerce and e-banking, owing to their high accuracy in real-world conditions. However, these systems are vulnerable to facial morphing attacks, which are generated by blending face color images of different subjects. This paper presents a new method for generating 3D face morphs from two bona fide point clouds. The proposed method first selects bona fide point clouds with neutral expressions. The two input point clouds were then registered using a Bayesian Coherent Point Drift (BCPD) without optimization, and the geometry and color of the registered point clouds were averaged to generate a face morphing point cloud. The proposed method generates 388 face-morphing point clouds from 200 bona fide subjects. The effectiveness of the method was demonstrated through extensive vulnerability experiments, achieving a Generalized Morphing Attack Potential (G-MAP) of 97.93%, which is superior to the existing state-of-the-art (SOTA) with a G-MAP of 81.61%.

3D Face Morphing Attack Generation using Non-Rigid Registration

TL;DR

The paper tackles morphing attacks on face recognition by introducing the first direct 3D morphing method that operates on 3D face point clouds. It registers two bona fide faces using Bayesian Coherent Point Drift (BCPD), then averages geometry and color to produce a morph, forming a three-step pipeline: 3D-3D alignment, colorization, and 3D morphing. Evaluated on the Facescape dataset with 200 identities, the method achieves high attack potential, with GMAP values up to 97.93% for 3D FRS and 100% for 2D FRS, surpassing the prior state-of-the-art and highlighting the need for robust 3D morphing detection. The work demonstrates the feasibility and strength of 3D morphing attacks and underscores important directions for improving FRS security against such threats.

Abstract

Face Recognition Systems (FRS) are widely used in commercial environments, such as e-commerce and e-banking, owing to their high accuracy in real-world conditions. However, these systems are vulnerable to facial morphing attacks, which are generated by blending face color images of different subjects. This paper presents a new method for generating 3D face morphs from two bona fide point clouds. The proposed method first selects bona fide point clouds with neutral expressions. The two input point clouds were then registered using a Bayesian Coherent Point Drift (BCPD) without optimization, and the geometry and color of the registered point clouds were averaged to generate a face morphing point cloud. The proposed method generates 388 face-morphing point clouds from 200 bona fide subjects. The effectiveness of the method was demonstrated through extensive vulnerability experiments, achieving a Generalized Morphing Attack Potential (G-MAP) of 97.93%, which is superior to the existing state-of-the-art (SOTA) with a G-MAP of 81.61%.
Paper Structure (11 sections, 4 equations, 2 figures, 1 table)

This paper contains 11 sections, 4 equations, 2 figures, 1 table.

Table of Contents

  1. Introduction & Related Work
  2. Proposed Method
  3. BCPD-based 3D-3D Alignment
  4. Notation
  5. Initialization
  6. Optimization
  7. Colorization of the aligned point clouds
  8. 3D Morphing Process
  9. Dataset Details
  10. Experiments & Results
  11. Conclusions & Future-Work

Figures (2)

  • Figure 1: Illustration showing block diagram of the proposed approach
  • Figure 2: Illustration showing Bona fide Input and Morphing Face Samples generated using proposed method and SOTA Singh_3DFaceMorphing_TBIOM23 where SOTA shows blending artifacts in facial boundaries.