This repository hosts the source code for my academic homepage: 👉 soumyaroopnandi.github.io
The website is built using Jekyll and the Minimal Mistakes theme, and automatically deployed via GitHub Pages.
I am a recent Ph.D. graduate in Computer Science from the University of Southern California (USC), specializing in Artificial Intelligence (AI), Computer Vision, and Deep Learning. My doctoral research at the USC Information Sciences Institute focused on developing state-space and information-theoretic deep learning models for AI-generated image forgery detection, generative model fingerprinting, and vision-language understanding.
I bring hands-on experience across the full AI stack — from model design and optimization (SSMs, VLMs, Diffusion Models, GANs, CNNs, Transformers) to data curation, multimodal learning, and large-scale deployment. My projects bridge vision, language, and generative modeling, with applications in biomedical imaging, digital content integrity, and trustworthy AI systems.
Over the years, I have gained extensive experience in AI-driven image analysis, generative model forensics, and visual-language alignment, spanning forgery detection, diffusion modeling, self-supervised learning, and anomaly detection. My research integrates semantic understanding with low-level artifact reasoning to advance explainable and robust visual intelligence.
Beyond academia, I have industry experience at ABB Robotics and ON Semiconductor, applying computer vision, 3D mapping, and reinforcement learning to real-world automation systems. I also worked at the Kansas Geological Survey and the Center for Remote Sensing of Ice Sheets, developing signal and image processing pipelines for radar and seismic data.
During my Master’s in Electrical Engineering at the University of Kansas, my thesis — “Robust Object Tracking and Adaptive Detection for Auto Navigation of Unmanned Aerial Vehicles” — explored object tracking, motion estimation, and adaptive detection using classical computer vision and signal processing techniques. This research formed the foundation for my later work in deep learning and generative AI.
With a strong foundation in AI model development, generative systems, and applied computer vision, I am passionate about building scalable, reliable, and interpretable AI solutions that drive innovation in visual intelligence and generative trustworthiness.
- AI-generated Image Forgery Detection
- Generative Model Fingerprinting
- Vision-Language Modeling
- Diffusion and State-Space Models
- Multimodal and Representation Learning
- Trustworthy and Explainable AI