Human Authentication Using Facial Cues

Human Authentication Using Facial Cues

by Fernando Alonso-Fernandez

Applied Sciences & Engineering AI — Models & Research UYQArtificial intelligence (AI)UYQVComputer vision
🎙 Fernando Alonso-Fernandez 👥 2K 📅 January 4, 2019 ⏱ 40 min 👁 52 📄 expert opinion 🧭 2026-08-18
Available in: English (current) Français

Keywords

biometricsface recognitionperiocular regionsoft biometricsauthentication

Summary

This plenary talk by Dr. Fernando Alonso-Fernandez from Halmstad University, presented at the Artificial Intelligence International Conference in Barcelona in 2018, discusses the use of facial cues for human authentication. The speaker introduces the concept of biometrics and its advantages over traditional tokens like passwords or ID cards. He emphasizes the importance of the periocular region (the area around the eye) as a biometric modality that can be used when the whole face or iris is not available or of sufficient quality. The talk covers various aspects: feature extraction methods for periocular recognition, comparison with iris and face recognition, expression estimation from the periocular region, eye detection without full face, cross-sensor recognition challenges, and the use of soft biometrics (age, gender, etc.) to complement hard biometrics. The speaker presents results from his research, showing that periocular recognition can achieve good performance, especially with visible light images, and that combining soft biometrics with face recognition can improve accuracy. He also discusses the challenges of unconstrained sensing and cross-sensor variability. The talk concludes with the potential of these methods for real-world applications, such as smartphone authentication and surveillance.

186 words

Critical Evaluation

Value of the Information & Strength of the Argument

The talk provides a comprehensive overview of the state of the art in periocular biometrics, highlighting its potential as a complementary or alternative modality to face and iris recognition. The speaker supports his arguments with experimental results from his own research and references to existing literature. He effectively demonstrates the advantages of the periocular region, such as its availability at various distances and its tolerance to occlusion. The argumentation is solid, though the presentation is more of a survey than a deep dive into any single study. The speaker also addresses practical challenges like cross-sensor performance degradation, which is a valuable contribution to the field.

Scientific Rigor, Source Quality, Title Accuracy

The speaker is an established researcher in the field, and the content is based on published research, as indicated by references to journal articles and book chapters. The talk is well-structured and the title accurately reflects the content. However, as a conference presentation, it lacks the rigor of a peer-reviewed paper, and some claims are presented without detailed evidence. The speaker mentions that the content is available in two published articles, which adds credibility. The title is appropriate and not misleading.

201 words

Title / Content Match

The title accurately reflects the content, which focuses on facial cues for authentication, particularly the periocular region.

Quality & Reliability

7/10

The talk is given by an academic expert in biometrics, presenting research findings and referencing published work. However, it is a conference presentation with limited depth and no formal peer review in this format.

Key Moments

Markers derived by PSI from the transcript: the creator did not define chapters.

Cited Sources

  • Periocular biometrics: a survey — Mentioned as a published article containing the content of the presentation.
  • Book chapter on periocular biometrics — Mentioned as a published book chapter containing the content of the presentation.

Concurring Sources

  • Periocular biometrics: a survey — The speaker's own publication, which supports the content.

Contribution & Novelties

The talk provides an overview of the speaker’s research on periocular biometrics, highlighting its potential as a robust biometric modality. It contributes to the field by summarizing recent advances and presenting new results on expression estimation and cross-sensor recognition. The speaker also discusses the integration of soft biometrics to improve recognition accuracy.

Pour aller plus loin :

  • Periocular biometrics — Overview of the modality.
  • Biometrics — General introduction to biometric recognition.
  • Face recognition — Related technology.

76 words

Radar Profile

The radar profile shows a balanced performance across all dimensions, with slightly higher scores in information quantity and quality, indicating a well-rounded presentation. The technical level is moderate, making it accessible to a broad audience while still providing depth.

Reliability 7/10