🛂 ScoreFusionABC

Score Fusion in Multimodal Automated Border Control Systems

MATLAB source code for the BIOSIG 2016 paper
Enhancing the performance of multimodal Automated Border Control systems

---

🧠 Overview

ScoreFusionABC provides a MATLAB implementation of score-level fusion methods for multimodal biometric verification in the context of Automated Border Control (ABC) systems.

The code is designed to combine biometric matcher outputs and quality information from multiple modalities in order to improve verification performance in border-control scenarios.

Typical modalities in multimodal ABC pipelines may include face, fingerprint, iris, or other biometric traits. This repository focuses on the fusion stage, assuming that biometric scores and quality measures are already computed by external biometric systems.

---

🇪🇺 ABC4EU Context

This repository is related to the ABC4EU European project, which investigated next-generation Automated Border Control technologies and multimodal biometric processing.

Project page:

http://iebil.di.unimi.it/projects/abc4eu

---

🔬 Processing Pipeline

flowchart LR
    A[Biometric Matchers] --> B[DATA_scores]
    C[Quality Estimators] --> D[DATA_qualities]
    B --> E[Score Normalization]
    D --> F[Quality Processing]
    E --> G[Score Fusion]
    F --> G
    G --> H[Verification Scores]
    H --> I[ROC / DET / EER Evaluation]

Loading

The repository implements a complete experimental workflow for:

• loading biometric scores and quality indicators,
• preparing genuine and impostor comparisons,
• applying score normalization and fusion strategies,
• estimating performance using biometric verification metrics,
• generating plots and statistics for analysis.

---

📁 Repository Structure

ScoreFusionABC/
│
├── launch_scoreFusionABC.m     # Main MATLAB entry point
│
├── DATA_scores/                # Input biometric scores
├── DATA_qualities/             # Input biometric quality measures
│
├── biometricUtil/              # Biometric utility functions
├── calcoloROC/                 # ROC / DET / error-rate computation utilities
├── fusions/                    # Score fusion algorithms
├── mixturecode2/               # Finite mixture model utilities
├── mLib/                       # Supporting MATLAB library functions
├── util/                       # General-purpose helper functions
│
├── LICENSE                     # GPL-3.0 license
└── README.md

---

🚀 Getting Started

1. Clone the repository

git clone https://github.com/AngeloUNIMI/ScoreFusionABC.git
cd ScoreFusionABC

2. Prepare the input data

The code expects pre-computed biometric scores and quality values in the following folders:

./DATA_scores/
./DATA_qualities/

These values must be generated by external biometric matchers or quality-estimation software.

See the .dat files in the repository for the expected input format.

3. Run the main script

Open MATLAB from the repository root and run:

launch_scoreFusionABC

---

📊 Outputs

The framework supports analysis of biometric verification performance through metrics and plots such as:

Output	Description
Genuine/impostor scores	Score distributions for biometric comparisons
ROC curves	Receiver Operating Characteristic analysis
DET curves	Detection Error Tradeoff visualization
EER	Equal Error Rate
Fusion scores	Combined multimodal verification scores
Quality-aware analysis	Use of biometric quality information in fusion

---

🧩 Implemented / Referenced Fusion Ideas

The repository includes code and experimental routines related to several score fusion strategies, including methods inspired by:

• likelihood-ratio-based biometric score fusion,
• quality-aware score fusion,
• finite mixture models,
• kernel Fisher discriminant analysis,
• weighted score combinations for multibiometric systems.

---

📚 Related Methods and Dependencies

Part of the code uses or refers to the following works and libraries.

Finite Mixture Models

M. Figueiredo and A. K. Jain,
“Unsupervised learning of finite mixture models,”
IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 24, no. 3, pp. 381–396, 2002.

http://www.lx.it.pt/~mtf/
http://www.lx.it.pt/~mtf/mixturecode2.zip

VLFeat

A. Vedaldi and B. Fulkerson,
“VLFeat: An Open and Portable Library of Computer Vision Algorithms,” 2008.

http://www.vlfeat.org/

Biometric Score Fusion References

• K. Nandakumar, Y. Chen, S. Dass, and A. Jain,
  “Likelihood ratio-based biometric score fusion,”
  IEEE Transactions on Pattern Analysis and Machine Intelligence, 2008.

• S. Mika, G. Rätsch, J. Weston, B. Schölkopf, and K. R. Müller,
  “Fisher discriminant analysis with kernels,”
  Neural Networks for Signal Processing IX, 1999.

• C. Chia, N. Sherkat, and L. Nolle,
  “Towards a best linear combination for multimodal biometric fusion,”
  ICPR, 2010.

• N. Damer, A. Opel, and A. Nouak,
  “Biometric source weighting in multi-biometric fusion: towards a generalized and robust solution,”
  EUSIPCO, 2014.

---

📖 Paper

If you use this repository, please cite:

@InProceedings{biosig16,
  author    = {A. Anand and R. Donida Labati and A. Genovese and E. Muñoz and V. Piuri and F. Scotti and G. Sforza},
  title     = {Enhancing the performance of multimodal Automated Border Control systems},
  booktitle = {Proc. of the 15th Int. Conf. of the Biometrics Special Interest Group (BIOSIG 2016)},
  address   = {Darmstadt, Germany},
  pages     = {1--5},
  month     = {September},
  year      = {2016},
  doi       = {10.1109/BIOSIG.2016.7736922}
}

Paper:

https://ieeexplore.ieee.org/document/7736922

---

👥 Authors

A. Anand, R. Donida Labati, A. Genovese, E. Muñoz, V. Piuri, F. Scotti, and G. Sforza

Department of Computer Science
Università degli Studi di Milano, Italy

---

📄 License

This project is released under the GNU General Public License v3.0.

See the LICENSE file for details.