Research Philosophy & Focus Areas
Investigating the confluence of real-time physical simulation, deterministic robotics kinematics, and neural visual representations.
Research Mission
Modern physical systems—from aerospace logistics airbases to marine submersibles and biological neural connectomes—demand computational models that do not merely visualize, but maintain deterministic mathematical fidelity.
My research investigates how browser-native spatial computing, CUDA-accelerated neural rendering, and graph-theoretic state machines can be coupled to bridge empirical telemetry with predictive simulation.
Core Academic Interests
High-Fidelity Physical Simulation in WebGL & Distributed Graphics
Developing browser-native 60 FPS simulations of large-scale automated campuses (e.g., WareTwin). Investigating multi-agent reservation grid algorithms, deterministic A* time-expanded routing, and low-latency state synchronization with hardware controllers via WebSockets and PX4/MAVLink protocols.
3D Gaussian Splatting & Real-Time Radiance Fields
Researching rasterization efficiencies in 3D Gaussian Splatting (3DGS). Investigating dual-pipeline rendering with CUDA acceleration and CPU depth-sorted projection fallbacks for heterogeneous client devices, alongside higher-order Spherical Harmonics color representations.
Multimodal Neuroimaging Twins & White-Matter Tractography
Constructing computational models from raw 4D DWI, fMRI, and anatomical MRI scans (NEUROTWIN). Evaluating microstructural biomarkers (Fractional Anisotropy, Mean Diffusivity) and modeling topological graph disruption caused by focal lesions using 59×59 connectivity matrices.
Manuscripts & Preprints
Active Research Under Preparation & Peer Review
In adherence to personal data integrity guidelines, citations, DOI links, and arXiv preprints will be indexed here once formally accepted or released by the author.
To add publications: update the publications array in src/config/profile.config.ts.
Research Implementations
NEUROTWIN: Multimodal Neuroimaging Digital Twin Laboratory
Transforms raw MRI, fMRI, and DTI neuroimaging data into an interactive computational digital twin representing cortical surfaces, 3D white-matter tractography, and structural lesion simulations.
Read Research Case Study →3D Gaussian Splatting Desktop Engine
A desktop 3D Gaussian Splatting viewer supporting millions of Gaussians with an automated pre-flight CUDA diagnostic and graceful CPU depth-sorting fallback.
Read Research Case Study →Document Layout Intelligence & Neural OCR
Transforms raw documents into structured JSON with bilingual neural OCR, region semantic classification, SLANet table extraction, and multi-column reading order reconstruction.
Read Research Case Study →NeuroFormer-X: Multimodal Signal Modeling
Multimodal signal encoding and cognitive state prediction using deep transformer backbones, digital signal filtering, and latent space representations.
Read Research Case Study →MedAI-LLM-Chatbot
A conversational language model system adapted for medical query synthesis, clinical triage guidance, and patient telemetry analysis.
Read Research Case Study →Academic Collaboration & Graduate Admissions
Open to collaborative research inquiries, laboratory graduate positions, and co-authorship opportunities in autonomous systems, robotics, and physical simulation.
Initiate Research Inquiry