Mohammod
Tareq Aziz
Justice
Machine Learning & Deep Learning enthusiast. Exploring the frontiers of spatiotemporal modeling, zero-shot learning, and computer vision to solve complex real-world problems.
About Me
I am a Computer Science graduate from BRAC University, Dhaka, with a strong research-driven interest in Machine Learning, Deep Learning, and Computer Vision. My academic and project work focuses on developing intelligent, data-driven systems, particularly in areas such as video understanding, anomaly detection, and applied machine learning.
My academic training in Computer Science, combined with hands-on experience in research-oriented projects, has naturally drawn me toward exploring the intersection of deep learning architectures, representation learning, and real-world problem solving. My primary research interests span Zero-Shot and Self-Supervised Learning, Vision Transformers, Spatiotemporal Modeling, and Multi-Modal Learning, with applications in surveillance, human behavior analysis, and data-driven diagnostics.
My undergraduate thesis centers on context-aware zero-shot anomaly detection in surveillance videos, where I designed a dual-stream framework leveraging spatiotemporal transformers, predictive modeling, and vision-language alignment to detect unseen anomalous events without requiring labeled anomaly data.
In addition to deep learning research, I have worked on applied machine learning projects involving spectrophotometric data analysis and classical ML models, strengthening my understanding of data preprocessing, feature engineering, and experimental evaluation. I am proficient in Python-based ML and deep learning ecosystems, including PyTorch, TensorFlow, and common scientific computing libraries.