Biography
I am Fengcheng Yu, currently an M.S. student in Computer Science at the University of Southern California (USC), where I began my graduate studies in August 2025. Since 2025, I have been involved in research led by Dr. Ramesh Govindan (Northrop Grumman Chair in Engineering, Professor of Computer Science at the University of Southern California), focusing on mobile systems, particularly real-time mobile sensing and low-latency data processing for in-vehicle perception systems, with an emphasis on efficient parallel processing under resource-constrained environments.
Previously, I received my Bachelor of Engineering degree from Sun Yat-sen University in June 2025. From 2022 to 2023, I conducted research in the YCLab led by Professor Calvin Yu-Chian Chen (陈语谦) (Director of AI for Science (AI4S) at Peking University). From 2024 to 2025, I continued my research training in Professor Guang Tan (谭光)’s laboratory at Sun Yat-sen University.
📚 Education
- University of Southern California 2025-2027
- MS Student, MS in Computer Science (CS32); GPA: 3.85/4.0
- Sun Yat-sen University 2021-2025
- Undergraduate Student, Intelligence Science and Technology, Computer Science; GPA: 3.9/4.0, 3.9/5.0;
- Undergrad TA for Operating Systems and Advanced programming (C++ programming);
🔥 Awards
- 2021-2024: Outstanding Student Scholarship, SYSU (1st Prize in 2022 🏅, awarded consecutively 2022-2024)
- 2022-2024: Academic Competition Scholarship, SYSU (awarded consecutively 2023 and 2024)
- 2022: 1st Prize of Asia and Pacific Mathematical Contest in Modeling (national-level)
- 2023: 2nd Prize of China University Computer Competition - Group Programming Ladder Competition - Guangdong
- 2023: 2nd Prize in the “LanQiao” Cup National Software Competition (C/C++) Guangdong Division
📝 Publications
*** Paper Title
Chunyu Xia, Christina Suyong Shin, Fengcheng Yu, Chuan Li, Fan Bai, Harsha V. Madhyastha, Ramesh Govindan
- Keywords: mobile sensing, gaze localization, multi-camera perception, real-time systems
Deflickering Vision-Based Occupancy Networks through Lightweight Spatio-Temporal Correlation
Fengcheng Yu, Haoran Xu, Canming Xia, Ziyang Zong, Guang Tan
- Keywords: vision-based occupancy network, deflickering, spatio-temporal correlation, plug-and-play
HouseTune: Two-Stage Floorplan Generation with LLM Assistance
Ziyang Zong, Guanying Chen, Zhaohuan Zhan, Fengcheng Yu, Guang Tan
- Keywords: floorplan generation, large language model, diffusion model
Parkinson’s severity diagnosis explainable model based on 3D multi-head attention residual network
Jiehui Huang, Lishan Lin, Fengcheng Yu, Xuedong He, Wenhui Song, Jiaying Lin, Zhenchao Tang, Kang Yuan, Yucheng Li, Haofan Huang, Zhong Pei, Wenbiao Xian, Calvin Yu-Chian Chen
- Keywords: Parkinson’s disease, severity diagnosis, deep learning, facial video, interpretable model
🔧 Selected Projects
Edge FaaS Cold-Start Mitigation via EWMA + CUSUM Prediction USC CSCI599, Network System for Cloud Computing Course Project, Professor: Ramesh Govindan
A clean-slate, bare-metal C++ FaaS worker targeting cold-start in edge serverless. Uses OS-level Copy-on-Write process primitives to isolate the prediction algorithm from container/KVM noise. Strict Reactor / 64-thread DispatchPool separation; the EWMA + CUSUM predictor runs in O(1) time and memory.
HareMQ Simulation implementation of RabbitMQ core functions
A C++ message queue inspired by RabbitMQ. Supports direct/topic/header/fan-out exchanges, message persistence, high availability, and multiple concurrent clients.
Related technologies: C/C++11, Boost, Protobuf, Muduo, SQLite3, Gtest
Boost Search Engine A search engine implementation based on the Boost library
A specialized search engine over the Boost library documentation. Covers the full pipeline — data preprocessing, inverted index construction, query processing, ranked result presentation — with word-frequency scoring and dynamic index updates.
Related technologies: C/C++11, STL, Boost, jsoncpp, cppjieba, cpp-httplib, jQuery, ajax, inverted indexing, ranking algorithms
Load-balanced Online Judge System An online algorithm question-solving service, similar to LeetCode
A LeetCode-like online coding and evaluation platform — write, submit, and judge code in the browser with instant feedback. Includes Ace editor integration and a custom load balancer over distributed multi-process/multi-thread judges.
Related technologies: STL, Boost, cpp-httplib, ctemplate, jsoncpp, MySQL C Connector, Ace editor, ajax, load balancing
Google Tcmalloc Simulation Implementation
A simulation of Google’s tcmalloc — a high-concurrency memory pool replacing native malloc, designed for multi-core servers. Aggregates same-size objects into page-aligned spans to reduce fragmentation and avoid per-core lock contention.
Related technologies: STL, multi-threading, memory management, lock-free programming
Multi-pattern Logging System
A C++ logging system with multi-level logging, sync/async modes, and pluggable output sinks (console, file). Thread-safe by design, applying classic OOP patterns throughout.
Related technologies: design patterns (singleton, factory, builder, proxy), producer–consumer, multi-threading, GTest
Event-Driven Pipeline Communication System Framework
An epoll-based event-driven IPC framework with a modular structure for easy integration and extension across different environments.
Related technologies: STL, system programming, multiplexing, pipe communication
- Dual-Thread-Pool-Based-Pipeline-Communication-System-Framework
- This project implements a pipeline communication system framework based on dual thread pools from scratch, which is essentially communication between processes. The thread pool object code of this project is highly maintainable and highly readable. Thread pool objects can be easily reused in other projects
- Reactor-based-HyperWebServer
- The project manually implemented an epoll server based on red brick tree from scratch, and on this basis, encapsulated it into a reactor asynchronous IO multiplexing server. Due to the multiplexing feature, this network server can withstand high concurrency.
💻 Experience
- 2024.01 - 2024.02, Yunhong Information Technology Co., Ltd., Guangzhou, China.
- Virtualization R&D Engineer Intern in R&D Department
- Initiated and developed the “virtual machine probe” using Libvmi to monitor KVM architectures effectively. Enhanced virtual machine-to-host interactions through spice-protocol, enabling clipboard sharing and file dragging.
- Produced detailed documentation on new tools and protocols, significantly improving development and testing efficiency within the R&D department. Established a Confluence workspace to enhance collaborative efforts among a five-person team.
- 2024.07 - 2024.09, Lingzong Technology (Shenzhen) Co., Ltd., Shenzhen, China.
- R&D Center Intern
- Conducted in-depth research on advanced occupancy prediction technologies such as SurroundOcc and SelfOcc. Focused on understanding and experimenting with their core functionalities and advantages.
- Researched and explored the mechanism of “multi-head prediction” in occupancy prediction, adjusted the original model structure to enhance prediction accuracy.
- 2021.10-Present, CSDN (One of China’s Largest Tech Community)
- Backend Development Content Creator, Part-time
- Selected as a high-quality backend development content creator.
- Published 150+ technical blogs covering system design, distributed computing, and backend frameworks. Received over 390,000 visits.
- Built an engaged community with 8,000+ followers.
- Selected highly rated columns (Posts are in Chinese only):
🌐 Open Source & Community
- LeetsGoAgain: LeetCode problem-solving journal — solutions with written reasoning and notes, organized by topic and difficulty.
- rookie-cplusplus: STL deep-dive — handcrafted implementations of core containers (string, vector, list, map, unordered_map) and data structures (AVL tree, red-black tree, skip list), along with algorithm library internals.
- rookie-csharp: Reference notes and annotated examples covering C# language fundamentals, written during initial exploration of the language.
- Sequence-Alignment-Machine: USC CSCI570 final project. Compares standard O(mn)-space dynamic programming with Hirschberg’s O(m+n)-space divide-and-conquer algorithm for DNA/protein sequence alignment, with empirical analysis of the memory–runtime trade-off.
- NimbusBackup: A cloud backup system that automatically uploads files from a specified local directory to a remote server. Supports browser-based viewing and downloading with HTTP range-based resumable download, and implements a hot/cold file management strategy that compresses infrequently accessed files to reduce disk usage.
- STL-based-Deque-implementation: A from-scratch implementation of
std::dequefollowing the STL design, including the segmented buffer layout and iterator arithmetic. - Reactor-based-HyperWebServer: A high-concurrency web server built on a hand-rolled epoll reactor with red-black tree event management and async I/O multiplexing.
- Dual-Thread-Pool-Based-Pipeline-Communication-System-Framework: An inter-process pipeline communication framework built on dual thread pools, with RAII mutex wrappers, thread object encapsulation, and a reusable
thread_controlcore that decouples thread management from client/server logic. - Multiplexing-high-performance-IO-server: Implementations of all three I/O multiplexing models (select, poll, epoll) in C/C++, composable with HTTP/Web servers and other SystemV I/O models.
- NLP-PoetryModels: Experiments on CharRNN-based poem generation with LSTM, GRU, RNN variants, and attention/bidirectional extensions across Chinese, English, and Japanese corpora. Evaluated with Tsinghua BERT-CCPoem and human scoring.