Sports Player & Ball Tracking
Multi-object detection and tracking system for sports video analysis. Uses Faster R-CNN for frame-level detection and Deep SORT for persistent ID assignment across frames — enabling player trajectory extraction and ball tracking for performance analysis.
Project Highlights
Faster R-CNN detector fine-tuned on sports footage for accurate player and ball bounding boxes.
Deep SORT tracker maintains consistent IDs across occlusions and fast motion frames.
Pipeline supports both offline video processing and real-time stream input.
Trajectory data exportable for downstream performance analysis and coaching insights.
Explore More Work
Deep dive into other high-performance solutions.

Enterprise hybrid RAG and financial multi-agent translation engine bridging Vietnamese Accounting Standards (VAS) and International Financial Reporting Standards (IFRS). Automates General Ledger transformations, standard-by-standard comparative analysis, and double-entry adjustments with mathematical debit/credit balancing and legal citation grounding.

Production-oriented TypeScript monorepo for an automated Facebook Messenger conversational commerce agent. Features secure Meta webhook ingress, durable event queues via Redis, ordered/idempotent worker pipelines, optimistic conversation state, and relational product catalog search.

End-to-end real-time credit card fraud detection system. Ingests transaction streams via Apache Kafka, applies a trained Random Forest model through PySpark Structured Streaming, stores predictions in PostgreSQL, and exposes a live Streamlit dashboard for monitoring and analytics.

Full-stack personal finance co-pilot tailored for Vietnamese young professionals and developers. Features dynamic wealth allocation modeling, dollar-cost averaging (DCA) simulations, monthly AI financial check-ins, automated webhook spending tracking, and multi-provider LLM routing (Gemini, OpenAI, Anthropic).

Intelligent fashion recommendation system utilizing sequence mining techniques, multimodal deep learning, and social media behavioral analysis. Combines image and text representations with LSTM and Transformer architectures to deliver personalized trend-aware outfit suggestions.

End-to-end churn analytics solution on 7,032 telecom customer records. Combines EDA, ensemble ML modeling, and a Power BI dashboard to identify at-risk customers, quantify revenue impact, and surface actionable retention strategies.

Automated data integration tool that processes Excel Binary Workbooks (XLSB) and synchronizes financial model data with subsidiary summary files across multiple entities, built on a modular COM-interface architecture.

Production-ready CLI tool that splits Excel workbook SG&A summary sheets into separate files by Department or Project, preserving original formatting.
Multi-object detection and tracking system for sports video analysis. Uses Faster R-CNN for frame-level detection and Deep SORT for persistent ID assignment across frames — enabling player trajectory extraction and ball tracking for performance analysis.

Distributed stock market forecasting pipeline using Apache Spark for feature engineering and three model tiers — Linear Regression, LSTM (PyTorch), and a Hybrid ensemble. Delivers predictions through an interactive Streamlit dashboard with technical indicators and real-time alerts.

Multi-label toxic comment classifier using a hybrid CNN + LSTM deep learning architecture. Detects six toxicity categories — toxic, severe toxic, obscene, threat, insult, and identity hate — to automate content moderation at scale.

Binary classification project predicting whether a bank customer will subscribe to a term deposit, based on demographic, transaction, and marketing-campaign features.

Full-fledged Java ordering and logistics management application for pizza shops. Implements graph data structures and Dijkstra's Shortest Path algorithm to calculate optimal delivery routes, estimated delivery times, and accurate distance-based delivery charges.

Statistical methods project applying clustering and segmentation techniques to customer data to identify distinct customer groups.

Enterprise relational database design and management system for core banking operations. Enforces rigorous Entity-Relationship Diagrams (ERD), normalized 3NF schemas, transactional ACID compliance, stored procedures, and audit logging for accounts, branches, transactions, and loans.
