Master’s thesis by ” Olivia Khalil Oraha Dawood” Department of Computer Science

Discussion of the master’s thesis in the College of Computer Science and Mathematics -Department of Computer Science entitled: “Machine Learning Model for Kinship Detection Based on Voice”

supervised by Dr. Yusra Faisal Mohammad

The thesis addressed kinship verification based on voice signals by using machine learning and deep learning techniques to identify shared vocal characteristics among family members.

The study addressed the establishment of a new voice dataset collected from real families (Kin-FVs), and the extraction of complementary speech representations using Colored Log-Mel Spectrograms and Self-Supervised Learning (SSL), along with the development and comparison of different models and fusion strategies for kinship verification.

The study aimed to develop an intelligent and effective system capable of verifying parent-child relationships using voice only, and to investigate the effectiveness of voice signals and their different representations in distinguishing kinship relationships.

This thesis contributes to the following Sustainable Development Goals:

1. Goal 9: Industry, Innovation and Infrastructure (through supporting research and innovation in artificial intelligence and biometric recognition).
2. Goal 11: Sustainable Cities and Communities (through contributing to the development of intelligent verification and recognition systems that can be employed in security and identity verification applications within smart communities).
3. Goal 16: Peace, Justice and Strong Institutions (through supporting forensic investigation and identity verification applications using voice-based biometric verification technologies).

The discussion committee consists of

Dr.   Ramadan Mahmood Ramo (Chairman)

Dr.  Zeena Nabil Jamil (Member)

Dr. Amera Istiqlal Badran (Member)

Dr. Yusra Faisal Mohammad (Member and Supervisor)