Aman Raj

I am a computer science and data-science-focused undergraduate pursuing research in machine learning, computer vision, and transformer-based models. I have interned on projects involving video-based accident prediction and image reconstruction using transform methods.

My work emphasizes designing novel architectures and metric improvements — for example, developing a Graph(Graph) framework that combined GCNs, GATs, and Transformers and improving adjacency definitions for better accident anticipation performance.

I have applied deep learning techniques across domains including medical imaging, federated learning for privacy-preserving video analytics, and protein sequence-to-structure modeling using large language models. I complement technical work with strong analytical, communication, and teamwork skills.

I seek opportunities to contribute to research-driven teams solving real-world problems in computer vision, deep learning, and applied optimization, bringing a track record of internships, academic projects, and competitive achievements.