Li Jiaxuan

Electronic Engineer & AI Researcher
South-Central Minzu University
+86 18971439511
1964128026@qq.com
Shenyang, China
Age: 22

Personal Profile

Passionate electronic engineer and AI researcher with expertise in embedded systems, computer vision, and deep learning. Proven track record in academic research, technical competitions, and entrepreneurial ventures.

Proficient in embedded development (STM32), PyTorch framework, YOLO series models, and deep learning algorithms. Experienced with feature engineering and mathematical modeling methods like LightGBM/XGBoost. Skilled in C/Python development with full project implementation capabilities.

Possess excellent logical analysis and leadership skills. Proactive, self-driven, and enjoy interpersonal communication. Energetic and capable of working efficiently in fast-paced environments.

Education

South-Central Minzu University
Sep 2020 - Jun 2024

Bachelor of Engineering in Electronic Information Engineering

Core Courses: Machine Learning, Digital Image Processing, Digital Signal Processing, Communication Principles, Signals and Systems

Work Experience

Xfusion Digital Tech | Customer Manager, Shanghai (ICT)
Jun 2024 - Nov 2024
  • Drove sales for server products, aligning features with market needs through cross-functional collaboration with R&D
  • Led enterprise bidding projects (e.g., China Eastern Airlines, COMAC) managing solutions and requirement tracking to ensure project success
  • Supported channel partners in market expansion, executing strategies to boost brand presence
  • Achieved ¥3M+ project sales during 3-month probation; named Best Trainee among campus hires
Xingyu Network Tech | Head, Taihang Studio (EdTech)
Mar 2024 - Jun 2024
  • Built full business system for academic tutoring (product/channel/marketing), ensuring seamless execution
  • Launched new business line achieving 100K+ RMB weekly revenue through precise positioning and innovative strategies
  • Developed marketing system & campus agent channels, boosting conversions via public+private domain model

Research Experience

National Natural Science Foundation of China Project (62201621) | Project Leader
Sep 2022 - Jun 2024

Project Topic: Intelligent Dual-Spectrum Detection System for Urban Low-Altitude UAV Swarms

Introduction & Challenges:

"Low-Small-Moving" (LSM) UAVs pose critical security threats in urban airspace. Existing detectors suffer from low accuracy (e.g., 80.7% mAP for YOLOv7) due to small target size (0.04%-0.12% image area), complex backgrounds, and environmental interference.

Proposed Method:

Two-Stage Framework: ViBe → Target Cropping → Enhanced YOLOv7 → Coordinate Regression

  • Stage 1: ViBe Preprocessing - Extract moving targets via background subtraction, then crop 640×640 regions centered on UAVs
  • Stage 2: Enhanced YOLOv7
    • ELAN-CA: Integrate Coordinate Attention (CA) into ELAN modules to capture spatial-channel features
    • SPDConv: Replace strided convolutions with Space-to-Depth layers to preserve fine-grained details
    • Target Regression: Map detected coordinates to original images
Experiments & Results:

Dataset: 8,249 images (Simple/Medium/Complex backgrounds)

Key Results: mAP@0.5: 96.5% (vs. 80.7% for YOLOv7)

Ablation Study: +ViBe: +8.9% → +CA: +3.1% → +SPDConv: +3.0%

Speed: 119.5 GFLOPs (Real-time on NVIDIA A100)

Visual Comparison: Higher confidence in complex scenes

Conclusion: Our method boosts LSM-UAV detection by 15.8% mAP via ViBe-based motion target focusing, CA-enhanced feature extraction, and SPDConv for small-object detail retention.

Publication: Cui, Y.Q., Li, J.X.*, Hou, L.G., et al. Enhanced YOLOv7 Method for Small UAV Target Detection in Urban Environments [J]. Computer Engineering and Applications (indexed in EI Compendex / Chinese Core Journal), 2024, 60(10): 237-245. (with supervisor as first author)

Competition Experience

2023 National Undergraduate Electronic Design Contest
2023

Achievement: National Second Prize

Team Role: Team Leader

Topic: Problem H Signal Separation Device

  • Built the core signal processing system based on an STM32 microcontroller, integrating an AD9959 module for sweep signal output
  • Processed signals via multiplier mixing synthesis, optimized signal-to-noise ratio through low-pass filtering and amplification circuits, then sampled via ADC
  • Designed a dynamic error correction algorithm to eliminate 10Hz frequency offset
  • Combined equal precision measurement method to build a closed-loop correction system, utilizing a 0.0001Hz precision DDS for signal reconstruction
  • Employed FFT algorithm for spectrum analysis to accurately identify AM/FM modulation types and parameters (ma/mf)
2022 Hubei Provincial Undergraduate Electronic Design Contest
2022

Achievement: Provincial Second Prize

Team Role: Team Leader

Topic: Problem H Signal Separation Device

  • Developed a signal separation device based on an STM32 signal processing architecture
  • Achieved dual-channel mixed signal acquisition using on-chip ADC and restored original signal waveform parameters via FFT frequency domain analysis
  • Innovatively employed an ADC/DAC hybrid architecture to complete the full signal separation-reconstruction process
  • Enhanced signal features using a multiplier array, resulting in output signal frequency error approaching zero
  • Designed a touchscreen interface for real-time display of demodulation parameters and synchronized DAC output of demodulated waveforms to an oscilloscope for verification
Other Competitions
  • 2023 China International College Students Innovation Competition Provincial Bronze Medal
  • 2022 8th China International "Internet Plus" Innovation and Entrepreneurship Competition Hubei Provincial Bronze Award
  • 2023 China Collegiate Computing Contest South-Central Regional First Prize
  • 2021 South-Central Minzu University Ethnic Dance Competition Silver Medal Award
  • 2022 South-Central Minzu University Ethnic Dance Competition Bronze Medal Award

Entrepreneurial Experience

Mengyang (Hubei) E-commerce | Partner (Apparel)
Nov 2024 - Present

Role & Company Overview: Partner at Mengyang (Hubei) E-commerce Co., Ltd., specializing in apparel e-commerce, responsible for the Douyin MOONYOUNG store operations since November 2024. Focused on building and managing the end-to-end commercial ecosystem.

MOONYOUNG Brand Concept: Pursuing genderless, undefined clothing.

Major Achievements: Achieved a monthly GMV exceeding 100,000 RMB within the first three months, transforming the store from an unknown brand with zero followers to a Douyin Silver Medal store with a 100% positive feedback rate.

Key Responsibilities:

  • Launched and managed Douyin MOONYOUNG store operations (products, channels, livestream) from scratch, overseeing daily business execution
  • Co-led apparel design, supply chain setup, and platform operations to drive project implementation
  • Managed supply chain clothing manufacturing and channel expansion contracts
  • Handled daily store management and live streaming operations

Campus Experience

Association Management

Served as Vice President of the Electronic Design Association at South-Central Minzu University. Responsible for organizing technical training, skill instruction, and project progress management for the association and related professional competitions.

Organized and executed:

  • The 2022 Truth-Seeking Cup Electronic Design Competition
  • The 2023 Truth-Seeking Cup Electronic Design Competition
  • The 2024 Harmony Cup Electronic Design Competition co-hosted with Kashgar University
Debating & Dance

Debating Career: Served as Member of the Debate Team at South-Central Minzu University. Participated in departmental and university-level debate competitions.

Achievements:

  • Champion, School of Electronic Information Engineering Freshman Tournament (2020)
  • Champion, South-Central Minzu University Qicheng Cup Tournament (2021)
  • Runner-up, South-Central Minzu University Communist Youth League