CV

Education

• 2024 - 2026

  Riverside, CA

  M.S. in Computer Science

  University of California, Riverside (UCR)

  • GPA: 3.88
  • Courses: Algorithms, Machine Learning, Robot Navigation, Large Language Models

• 2020 - 2024

  Shenzhen, China

  B.E. in Computer Science and Technology

  Southern University of Science and Technology (SUSTech)

  • Courses: Data Structures and Algorithms, Database Principles, Object-Oriented Design, Software Engineering, Distributed Systems

Research Experience

• 2025 - 2026

  Riverside, CA

  Graduate Research Assistant

  UCR TASL Lab

  • Developed a vehicle–drone collaborative simulation framework using CARLA, SUMO, AirSim, and ROS 2, and extended a 7B vision-language-action model for aerial Vision-and-Language Navigation (VLN)
  • Built an automated data pipeline that generated the SAR-Drone-VLN-3K dataset with 3,000 search-and-return trajectories and 66K image–action samples
  • Fine-tuned the navigation policy using LoRA (PEFT) in PyTorch on 8×RTX 6000 Ada GPUs, reducing validation loss by 4× and improving action accuracy 2.5× (14% → 38%)

• 2022 - 2024

  Shenzhen, China

  Undergraduate Research Assistant

  SUSTech GameAI Lab

  • Built a configurable multi-agent pursuit-evasion simulation platform based on OpenAI Multi-Agent Particle Environment (MPE), supporting cooperative reinforcement learning with customizable environments
  • Implemented and trained multi-agent reinforcement learning (MARL) algorithms, including MADDPG, QMIX, and IQL, using Python and PyTorch across 25+ experiments and 50,000 training episodes, proving a GRU-based architecture as the best accuracy-efficiency trade-off and achieving an 80%+ capture success rate

Industry Experience

• 2023 - 2024

  Shenzhen, China

  Software Engineer, Intern

  Mushroom Wealth Technology

  • Built a news center subsystem with Vue.js and REST APIs, supporting 500+ users across 2 securities firms
  • Integrated 37 REST APIs and 9 lazy-loaded routes, reducing average data-loading latency from 3.7s to <0.5s
  • Extended a configuration-driven permission framework with feature flags for client-specific workflows, eliminating customer-specific code branches and improving maintainability
  • Migrated content from single-select to multi-select across authoring, review, and publishing workflows with backward compatibility; achieved 95% test coverage and contributed to CI/CD

• 2021 - 2021

  Beijing, China

  Computer Vision Algorithm Engineer, Intern

  Beijing CELinkout Information Technology

  • Fine-tuned YOLOv5 object detection models to identify surface defects on metal components, achieving 0.87 mAP@0.5 across 6 defect classes on a 3K-image dataset
  • Developed a detection and classification pipeline for maritime vessels in satellite imagery, reaching 98%+ classification accuracy

Projects

• Visiting SUSTech: WeChat Mini Program for Campus Access Control

  Full-Stack Developer

  • Developed and deployed a full-stack WeChat Mini Program serving 10,000+ users over 6 months, enabling appointment booking, QR-code verification, and status tracking
  • Implemented role-based appointment workflows using Spring Boot, REST APIs, and PostgreSQL, supporting user submission, admin review, and approval/rejection state transitions
  • Developed frontend components using WeChat Mini Program framework and Vue.js, and containerized the application with Docker for deployment

• Scanning Tunneling Microscope (STM) Image Analysis

  Computer Vision

  • Built an STM image dataset of ancient bronze-mirror clay molds, developing Python-based pipelines for image preprocessing, annotation, and data preparation
  • Developed an end-to-end computer vision pipeline using SAM and VGG-based models for particle segmentation, classification, and quantitative analysis, supporting archaeological research at SUSTech

Skills

Teaching

• 2025 - present

  Riverside, CA

  Reader for CS218 Algorithms @ UC Riverside

  • Spring 2025, Spring 2026