Lokesh Poudyal

At Lovely Professional University, I’m completing an MSc in Data Science in collaboration with IBM, building on my BSc in IT with a Cybersecurity specialization. My studies and training have included machine learning pipelines, advanced analytics, and IBM Cloud environments.

I’m developing a Medical Deepfake Defense Framework to detect synthetic tampering in CT and MRI scans. The project uses deep learning and frequency-domain feature extraction to identify anomalies and generate probabilistic risk scores for scan integrity.

I also developed an exam cheating detection system using Xception and InceptionV3 with attention mechanisms, achieving over 90% accuracy on primary classes. I presented its findings at the RACS-2025 conference, and built a Python tool that analyzes Android packages for cloning and malware risks.