Portail des thèses et publications scientifiques du Citadel
This study evaluates acoustic parameters for detecting deepfake audio signals, analyzing spectral and temporal features to distinguish authentic from synthetic speech. Results show high accuracy using MFCC and pitch variation metrics.
📄 PDFPresents an automated pipeline converting unstructured text into OWL ontologies using NLP and transformer-based models. Evaluated on domain-specific corpora with promising precision and recall scores.
📄 PDFProposes a speech recognition system adapted to the Moore language (spoken in Burkina Faso), addressing data scarcity with transfer learning and data augmentation techniques. Achieves competitive word error rates.
📄 PDFIntroduces a static and dynamic analysis framework for detecting unauthorized data exfiltration in Android apps, identifying sensitive API misuse and network leaks across a large app corpus.
📄 PDFA hemorrhagic stroke is a life-threatening medical condition that happens when a blood vessel in your brain ruptures and bleeds — it constitutes a burden on health services and the victim's family. The current definitive diagnosis of stroke is based on brain scanning and AI-assisted methods are explored to accelerate detection.
📄 PDFAxes de Recherche
