Lingyun Chen

About Me

I am a Master's student in Computer Engineering at NYU Tandon School of Engineering, with a strong foundation in Electronic Science and Technology from Beijing University of Technology. My expertise spans AI/Machine Learning, software development, and computer vision.

I'm passionate about developing AI-powered applications and solving complex technical challenges. With experience in Python, Java, C/C++, and various frameworks, I specialize in building scalable systems and implementing cutting-edge solutions in machine learning and AI.

Technical Skills

PyTorch Transformers Scikit-learn Federated Learning AIGC Computer Vision Python Java C/C++ SQL JavaScript Spring Boot Docker AWS React

Resume

Education

New York University

Master of Science in Computer Engineering

Sep 2023 - May 2025 (Expected)

University of Kentucky

Electrical Engineering

Dean's List Spring 2022

Aug 2021 - Jun 2022

Beijing University of Technology

Bachelor of Science in Electronic Science and Technology

Sep 2019 - Jul 2023

Work Experience

BioMap

Algorithm Intern, AI Engine R&D Department

Dec 2024 - Present

Contributed to developing the Xtrimo DNA pre-training model (BERT-based) by implementing the Masked Language Model (MLM) component
Benchmarked Xtrimo against other models and refined its architecture to enhance performance
Designed automated Python AI pipelines for DNA sequence data processing and model evaluation
Enhanced pipeline performance with caching strategies and Docker containerization

AdaSeco

AI Engineer, Customer Support Chatbot Development

Jul 2024 - Sep 2024

Developed a customer support chatbot using LLMs and Retrieval Augmented Generation (RAG)
Optimized chatbot response time, reducing average latency by 30%
Contributed to a 25% increase in customer satisfaction scores

Wissee Inc

AI Engineer, Graphic Generation Project

May 2024 - Aug 2024

Developed AIGC models (SDXL, ComfyUI) to generate clothing patterns based on social media trends
Built a pattern database with MySQL and integrated Elasticsearch for efficient semantic search
Conducted trend analysis on social media and e-commerce data to inform model training

Featured Projects

Speech Transcript Analysis and Summarization Framework

Developed a PyQt6 desktop application providing a graphical user interface for real-time speech transcript processing and generating summaries via various LLMs, with bilingual support and OCR-based text extraction.

Python PyQt6 LLMs OCR

Federated Learning with Imbalanced Data

Investigated the impact of non-IID data distributions on federated learning performance, developing simulated FL environments to evaluate resampling techniques for improving model accuracy and fairness metrics.

Federated Learning PyTorch CNN Data Resampling

Automatic Labyrinth Structure Localization

Applied YOLOv7 deep learning object detection model to automatically localize complex labyrinth structures within temporal bone CT scan images, with Python and OpenCV preprocessing pipelines.

YOLOv7 Computer Vision OpenCV Medical Imaging

Marine Chemical Pollution Detection

Constructed a dataset of 800 aerial ocean images and applied CNN models with transfer learning (VGG16) to classify ocean regions as polluted or normal, demonstrating the feasibility of ML for environmental monitoring.

MATLAB CNN VGG16 Transfer Learning

Publications

Implementing and Evaluating Simple Resampling Techniques in Federated Learning for Imbalanced Data

The 6th International Conference on Computing and Data Science (CONF-CDS 2024)

September 12, 2024. Portsmouth, UK

View Publication

Detection of Marine Chemical Pollution Based on Image Processing and Machine Learning

Proceedings of the 2022 6th International Conference on Electronic Information Technology and Computer Engineering (EITCE'22)