Curriculum Vitae

Bio

Building on my undergraduate degree in Electrical Engineering from SHU, I launched my career as an Embedded Software Engineer at JABIL, where I spent two years developing and optimizing software for advanced IoT and 4G/5G technologies. This foundational experience in hardware-software integration fueled my ambition to tackle more complex challenges at the intersection of connected devices and artificial intelligence.

In August 2021, I deliberately transitioned from industry to academia to pursue this goal, earning my M.Sc. in Electronic Engineering from CUHK. My graduate research and subsequent work have focused on solving cutting-edge problems in AI. My contributions to the field are highlighted by my first-author research on PSMGD, a novel algorithm for fast multi-objective optimization published at the AAAI conference, and my co-authored work on practical adversarial attacks on WiFi sensing, published at the MobiCom conference.

This research demonstrates my expertise across a range of high-impact domains, including multi-task learning, security, and privacy in machine learning. I have applied these skills to diverse challenges, such as developing vision-assisted optimization for medical lesion detection, as explored in my master’s thesis. I am passionate about leveraging my interdisciplinary background to build innovative solutions that bridge the gap between hardware and software intelligence 😜.

Education

Rochester Institute of Technology, Rochester

Jan 2024 - Jan 2025 Research Assistant in Computing and Information Sciences. GPA: 3.94/4.00.

Temple University, Philadelphia

Jan 2023 - Dec 2023 Teaching Assistant in Computer and Information Science. GPA: 3.89/4.00.
Teaching class: CIS 3515 (Introduction to Mobile Application Development).

The Chinese University of Hong Kong, Hong Kong

Aug 2021 - Nov 2022 Master of Science in Electronic Engineering. GPA: 3.275 (MERIT), the maximum GPA attainable is 4.000. Academic advisor: Prof. Hongliang Ren

Shanghai University, Shanghai

Sep 2015 - Jul 2019 Bachelor of Engineering in Electrical Engineering & Automation. GPA: 3.14/4.00 (83.3/100), rank: 31.4%. Academic advisor: Prof. Zhiyuan Gao

Honors & Certificates

• Fall 2016 National Encouragement Scholarship & Self-improvement Scholarship of SHU
• Fall 2017 National Encouragement Scholarship of SHU
• Spring 2018 First Prize in the Service Robot Special Competition of 2018 China Robot Competition
• Summer 2018 Second Prize in the TI Cup Undergraduate Shanghai Electronics Design Contest
• 12/2020 Course Certificate (OUTSTANDING) of Python Language Foundation and Application of PKU
• 01/2021 Certificate of Deep Learning Specialization (all 5 courses) of DeepLearning.AI in COURSERA
• 01/2021 Course Certificate of Machine Learning of Stanford in COURSERA
• 06/2021 Certificate of Python for Everybody Specialization (all 5 courses) of UMich in COURSERA
• 09/2022 Certificate of Google Data Analytics Specialization (all 8 courses) in COURSERA
• 11/2022 M.Sc. Certificate of Merit of CUHK Graduation Scholarships

Skills

1. Programming Skills: C/C++ (NCRE C Programming Level II, good), Matlab, LaTex, Python, Linux, Git, Kotlin, etc.
2. English Proficiency: IELTS Academic overall 7.0 (6.0); Duolingo overall 125.
3. Machine Learning: have a certain understanding of image processing (segmentation, classification, etc.), video analysis (object recogintion, detection, etc.).

Projects & Experiences

The NICT 2021 Embedded Dev & AI Summer School on Jul 2021

Keywords: Classification, TensorFlow, ART-Pi, RT-Thread.

1. Based on garbage datasets online, applied TensorFlow to get and expand images’ features, set the parameters such as structure, batch size, and then created sequential for training. Checked the model’s accuracy by matplotlib, then saved the model as a h5 file (tflite is also suitable).
2. Took AI-Tool to load the file and related configuration to ART-Pi, and then designed the driver codes under RT-Thread. After initialising the screen, camera, and so on, implemented the bilinear interpolation to convert captured images to three-channel data according to the model’s input. And relying on the predicted, there would be corresponding reactions on the ART-Pi.
3. Please refer to github.com/MarcuXu/Garbage-Classification-Device for more details 😘.

Design Engineer (Intern & Embedded Software) at Jabil Circuit (Shanghai) Co., Ltd from Mar 2019 to Jun 2021

Keywords: MCU, Circuit schematics, Bus communication protocols, Vivado, RRU, Driver design.

1. Responsible for estimation and application of MCU used in projects; assisted hardware engineers in testing temperature-controlled chips with Arduino R3, such as LMP92066.
2. Improved understanding of circuit schematics and using laboratory instruments; further gained to understand BSP and hardware layer of the MCU’s bus protocol such as IIC, SPI, DMA, UART, etc.
3. Gained the NMEA format and how to decode it in C; implemented the various commercial applications of RT-Thread; acquired to use HC-02 BT module, L80 GPS module, etc.
4. During design space exploration, developed transforming wdb files to vcd in Vivado by Python to process wave data to get its transition moment and value to analyze data per clock edge.
5. Completed the driver design of hardware chips on RRU, such as driver codes for CAT24M01, LM5064 and different instructions codes on hardware abstraction layer and other layers.
6. Gained to test Xilinx’s suite’s Processing System & Programming Logic with Vivado; acquired how to use XSCT and write TCL scripts to load the bootloader and other settings.
7. Acquired how to write JSON files, serialize and deserialize them with Python; acquired how to use Git and Azure DevOps in teamwork.

The TI Cup Undergraduate Shanghai Electronics Design Contest on Jul 2018

Keywords: Loop Current Signal Detection Device, STM32F1, RTOS, Driver design.

1. Led the team to design the Loop Current Signal Detection Device, which took STM32F1.
2. Assisted the selection of MCU and hardware components; responsible for applying RT-Thread to the MCU, and the driver design of interaction functions including the keypad, LCD and touchscreen, etc.
3. Responsible for the report writing, on-site installation and demonstration of the competition work, answered the questions about system design from the judges.

The Service Robot Special Competition of 2018 China Robot Competition on May 2018

Keywords: OpenMV, WeChat mini program, Auto-tracking, Bluetooth.

1. Led the team, helped to design a WeChat mini program to control the smart car. Based on MSP432, applied Bluetooth synchronization (latency < 0.1s) and MQTT (latency < 0.3s) to control the car and receive its real-time feedback, and realized auto-tracking by OpenMV.
2. Analyzed the competition topic with teammates, learned and selected modules such as camera, motors, etc.; developed the communication protocol with the host unit and assisted in debugging it.
3. Responsible for the software design of MCU and took FreeRTOS as a framework to drive sensors.

Theses & Publications

• M. Xu, P. Ju, J. Liu, and H. Yang, “PSMGD : Periodic stochastic multi-gradient descent for fast multi-objective optimization,” in Proceedings of the AAAI Conference on Artificial Intelligence, vol. 39, no. 20, 2025, pp. 21770–21778. (CCF-A, CORE-A*)
• Li, C., Xu, M., Du, Y., Liu, L., Shi, C., Wang, Y., Liu, H., & Chen, Y., “Practical Adversarial Attack on WiFi Sensing Through Unnoticeable Communication Packet Perturbation,” in Proceedings of the 30th Annual International Conference on Mobile Computing and Networking, 2024, pp. 373-387. (CCF-A, CORE-A*)
• Yuan, S., Liang, B., Wong, P. W., Xu, M., Li, C. H., Li, Z., & Ren, H., “Magnetic-Guided Flexible Origami Robot toward Long-Term Phototherapy of H. pylori in the Stomach,” in 2024 IEEE International Conference on Robotics and Automation (ICRA), 2024, pp. 9851-9857. (CCF-B, CORE-A*)
• Mingjing Xu, master’s thesis, “Modeling and vision-assisted optimization of sampling strategies for lesion detection”.

Interests

I often do sports training after school, like table tennis and badminton. I am also a big fan of Lionel Messi and FC Barcelona, so I usually spend weekends losing weight and meeting new people by attending football matches. I like to enjoy the natural scenery. My hometown has a piece of World Natural Heritage scenic spot, Mount Langshan. During my undergraduate and master’s vacations, I would ask my family and friends to go hiking there. Meanwhile, I have a passion for mobile games. I prefer to play as an Attack Damage Carry/Core (ADC, NOT the Analog to Digital Converter here 😅) in the LoL Mobile (Wild Rift) and stay between Master and Grandmaster in current ranks. Some match highlights are posted on my social media, and I’ve met many nice friends abroad through the game.