Speech & speaker recognition
Research into deep neural networks and speaker representations, including convVectors for speaker recognition.
Read the published paper ↗PhD, HDR — Associate Professor at École Supérieure de Technologie de Safi, Cadi Ayyad University, researching deep learning, generative AI and speech technology.
I'm Dr. Soufiane Hourri, an Associate Professor at École Supérieure de Technologie de Safi, part of Cadi Ayyad University in Morocco. I hold a PhD and an HDR (Habilitation à Diriger des Recherches) — a qualification recognizing research leadership and the ability to supervise doctoral candidates.
My research centers on deep learning and its applications to speech and language: automatic speaker recognition, convolutional neural network architectures for voice biometrics, generative AI, and neural machine translation for Arabic and other underrepresented languages. My publication record includes international, peer-reviewed journals. You can explore individual papers below or visit Google Scholar for the complete record and current citation metrics.
Beyond the lab and the classroom, I work with companies as an AI consultant — helping teams move from research-grade ideas to systems that run reliably in production. I believe the gap between academic AI and applied AI is smaller than most people think, and closing it is where I do my best work.
Teaching and leading research in deep learning, generative AI, and speech & language technologies. Supervising graduate research under HDR accreditation.
Awarded the qualification to independently direct doctoral research, recognizing an established, original body of published work.
Doctoral research focused on deep learning approaches to automatic speaker recognition, laying the foundation for my most-cited publications.
Research experience and academic exchange in the Paris region, contributing to international collaborative work.
Research into deep neural networks and speaker representations, including convVectors for speaker recognition.
Read the published paper ↗Co-authored research reviewing the challenges, approaches, and development of Arabic machine translation.
Explore the Arabic MT survey ↗Co-authored work on predicting MapReduce job performance in cloud environments using machine learning.
Read the open-access research ↗AI strategy, model development, generative AI integration, and workshops shaped around your team’s goals.
See how we can work together →