SR Logo/AI Engineering
PROJECTS CATALOG

Engineering & Research Systems

Complete index of machine learning architectures, spatial perception pipelines, multimodal agents, and high-throughput systems.

Showing 7 of 7 projects
PROJECT 01Computer Vision
PRODUCTION

Autonomous Vessel Tracking & Visual Re-Identification System

Production maritime perception pipeline integrating YOLO vessel detection, OC-SORT tracking, and LoFTR + DINOv2 feature descriptors deployed on live RTSP streams via FastAPI.

Technical Problem:

Real-time vessel tracking on live maritime RTSP video feeds suffers from false-positive track fragmentation, wake interference, wave backscatter, and severe visual appearance changes as vessels change bearing.

System Pipeline Preview
1.
Live RTSP Frame IngestionOpenCV / RTSP
2.
YOLO Vessel DetectionYOLO / PyTorch
3.
OC-SORT Motion-Gated TrackingOC-SORT / Kalman
4.
LoFTR + DINOv2 Feature MatchingLoFTR / DINOv2 / FastAPI
YOLOOC-SORTDINOv2LoFTRFastAPI+3
PROJECT 02Robotics / Sensors
PRODUCTION

PTZ Camera + LiDAR Multi-Domain Sensor Fusion Node

Spatial perception and tracking architecture integrating 3D LiDAR point clouds with PTZ optical cameras via ROS 2 and Extended Kalman Filtering.

Technical Problem:

Optical PTZ cameras provide detailed semantic classification but lack direct metric depth, while LiDAR point clouds offer accurate range but lack dense semantic labels; calibrating moving PTZ optical axes with LiDAR presents spatial-temporal synchronization challenges.

System Pipeline Preview
1.
Point Cloud PreprocessingPCL / ROS 2
2.
Dynamic PTZ Frame TransformationROS 2 tf2 / C++
3.
Spatial Sensor AssociationHomogeneous Transforms
4.
State Estimation & FilteringEKF / Eigen
ROS 2LiDAR (PCL)Extended Kalman FilterOpenCVC+++2
PROJECT 03Generative AI
COMPLETED

MedBot: AI Medical Knowledge Assistant & RAG Pipeline

Grounded retrieval-augmented generation pipeline indexing medical documents in Pinecone vector database with LangChain orchestration, OpenRouter, and FastAPI.

Technical Problem:

Medical Q&A systems require strictly factual, grounded information retrieval from complex medical literature; naive language models hallucinate terminology, dosages, and clinical guidelines without verifiable passage grounding.

System Pipeline Preview
1.
Medical PDF Ingestion & ChunkingPyPDF / LangChain
2.
Semantic Vector EmbeddingHugging Face / Pinecone
3.
Context Retrieval & FilteringPinecone Vector DB
4.
Grounded Response GenerationOpenRouter / FastAPI
LangChainPineconeFastAPIFlaskOpenRouter+2
PROJECT 04Computer Vision
PRODUCTION

Dual-EMA Real-Time Perimeter Intrusion Detection Pipeline

Classical computer vision perimeter security pipeline utilizing dual-EMA background subtraction, contour tracking, and global lighting change suppression on live CCTV feeds.

Technical Problem:

Perimeter security CCTV feeds experience continuous false alarms triggered by sudden cloud cover, wind-blown foliage, shadows, and camera auto-exposure adjustments when using standard motion detection.

System Pipeline Preview
1.
Stream Preprocessing & Grayscale NormalizationOpenCV / NumPy
2.
Dual-EMA Background ModelingDual-EMA Algorithm
3.
Global Lighting Shift SuppressionOpenCV Morphological Ops
4.
Contour Analysis & Intrusion AlertingPython / FastAPI
OpenCVClassical Computer VisionDual-EMAPythonFastAPI+1
PROJECT 05Generative AI
ONGOING

ASTRION: AI Logistics Support Agent

Source-grounded support agent for logistics operations that answers order, ticket and policy questions from SOPs and signed customer agreements, computes fees and credits in code, and changes nothing without confirmation.

Technical Problem:

Logistics support answers depend on order records, open tickets and contract terms, and figures such as cancellation fees and service credits must be exact. A model that writes those numbers itself, cites nothing, or changes records unprompted cannot be trusted or audited.

System Pipeline Preview
1.
Workspace-Scoped ContextWorkspace scoping / ASTRION API
2.
Cited Policy RetrievalRAG / Source citations
3.
Request-Time Record LookupAgentic tool use
4.
Deterministic Calculation & Confirmed ActionsCode-computed / Human-in-the-loop
RAGAgentic Tool UseSource CitationsHuman-in-the-LoopNext.js+1
PROJECT 06Computer Vision
COMPLETED

Maritime Low-Visibility Video Restoration Pipeline

Video enhancement and image restoration pipeline engineered to mitigate fog, heavy haze, and low-visibility conditions in naval surveillance operations.

Technical Problem:

Naval defense and maritime surveillance cameras operate under severe visual degradation—dense marine fog, sea spray, night glare, and low illumination—which drastically impedes target detection.

System Pipeline Preview
1.
Frame DecompositionOpenCV / NumPy
2.
Atmospheric Light & Fog EstimationDark Channel Prior / Python
3.
Deep Neural Edge RestorationPyTorch / CUDA
4.
Dynamic Range RecombinationCLAHE / OpenCV
PyTorchOpenCVCNNsPythonImage Enhancement+1
PROJECT 07Full-Stack AI
COMPLETED

SubSpace: AI-Powered Authentication & Chat Platform

Full-stack AI communication platform with secure session authentication, OpenRouter/Hugging Face model integration, and n8n workflow automation.

Technical Problem:

Modern web applications require seamless integration of AI capabilities with robust user session management, database security, and automated background workflows without latency bottlenecks.

System Pipeline Preview
1.
Responsive Frontend & AuthReact.js / TailwindCSS / Supabase
2.
Multi-Model AI RoutingOpenRouter / Hugging Face
3.
n8n Workflow Automationn8n / Webhooks
4.
Real-Time Data SynchronizationSupabase / PostgreSQL
React.jsViteTailwindCSSSupabasen8n+2