Peer-reviewed work in AI, speech & language
My research explores how deep learning and generative AI can understand and reproduce human speech and language — from speaker recognition to Arabic machine translation.
Publications, methods & collaboration
Explore selected work below. Each paper links to its publisher or a scholarly search; current citation counts are available on Google Scholar ↗.
Learning speaker identity from voice.
Our convVectors study explored CNN-based speaker representations under clean and noisy conditions. On the THUYG-20 SRE benchmark, the paper reports an equal error rate of 1.05%.
This is a result from the published experiment; performance depends on the dataset and evaluation conditions.
Read the Study ↗What I research
Speaker & Speech Recognition
Deep learning and CNN-based approaches to automatic speaker recognition — my most-cited line of research.
Generative AI & Deep Learning
Architectures and training strategies for generative and deep neural models applied to real-world data.
NLP & Machine Translation
Natural language processing with an emphasis on Arabic — a linguistically complex, lower-resource language.
Journal & conference papers
A selection of my published research. For the complete and continuously updated list, visit my Google Scholar profile.