About

About

I am a junior in school of Cyber Science and Engineering, Huazhong University of Science and Technology.

My Complete Resume: RESUME-ENRESUME-CN

I am enthusiastic about the prospect of furthering my studies in China, particularly in the field of efficient artificial intelligence. Throughout my undergraduate program, I have not only achieved excellent academic results, ranking 6th out of 98 with an average GPA of 4.43/5, but also actively involved in various research projects.

While I have a clear interest and some research foundation in the field of efficient AI, I am equally open and curious about unfamiliar research areas. I believe that through further study and continuous learning, I can make continuous progress on the path of scientific research and explore more possibilities.

I am looking forward to joining a research group that aligns with me and embarking on my graduate studies, climbing the academic summit together with like-minded peers.

Contact Information

Education

Huazhong University of Science and Technology (2021 - 2025) Bachelor in Cyberspace Security

GPA: 4.43/5.00, Major Ranking: 6/98, English Level: CET-6 557, CSP Certification: Top 7%

Courses: Calculus (94), C Language Programming (88), Discrete Mathematics (100), Data Structures (95), Assembly Language Programming (92), Computer Networks (92), Principles of Computer Organization (88), Principles of Compilers (91), Algorithm Design and Analysis (92), Comprehensive Course Design in Programming (96)

Awards

  • China Collegiate Programming Contest (CCPC) - Bronze Medal (2023)
  • National Undergraduate Mathematical Modeling Competition, Hubei - Third Prize (2023)
  • Software Category, Lanqiao Cup Provincial Contest - Second Prize (2023)
  • National English Competition for College Students, Provincial Level - Third Prize (2023)
  • MathorCup University Mathematical Modeling Challenge - Bronze Medal (2023)
  • China Collegiate Programming Contest (CCPC) - First Prize (2023)
  • Huawei ICT Competition, Ascend AI Track, National Level - First Prize (2024)

Experience

Innovative Training Project — Predicting Chip Test Termination Based on Diffusion Model

2023.03 - 2023.11
Investigated the application of diffusion model algorithms in the EDA chip domain, improving fault diagnosis efficiency and quality by transforming test feedback into images. Designed machine learning algorithms combined with dynamic test termination prediction, using the Denoising Diffusion Implicit Model to expand the image dataset for training.

  • Achieved higher accuracy and nearly 90% cost savings in test cost with the new scheme compared to the original one.
  • Team contributed the paper “Translating Test Responses to Images for Test-termination Prediction via Multiple Machine Learning Strategies,” submitted to ACM-TODEAS (CCF-B).

HUST-University of Sydney Summer School — Parallel Programming Practice

2023.07 - 2023.08
Learned basic parallel programming methods and implemented performance optimization for shared and distributed memory parallel computing systems using OpenMP or MPI interfaces, and CUDA programming optimization for GPU design.

On-Campus Research — Federated Learning for Medical Image Segmentation

2023.12 - Present
Researched the application of federated learning in medical image segmentation, addressing issues of complex model structure and high computational cost.

  • Explored and integrated Quantization Aware Training (QAT) for model compression and inference acceleration, reducing computational load to approximately one-fourth of the floating-point precision with minimal loss of accuracy.
  • The team is currently writing a paper expected to be submitted to IEEE-BIBM (CCF-B).

Information Security Competition — Multimodal Audio Attack Detection Based on Neural Networks

2023.12 - Present
Designed and implemented an intelligent detection system for multimodal audio attacks in voice recognition systems, showing good performance in detecting synthetic, replay, and adversarial attacks.

  • Developed an efficient voiceprint feature extraction and analysis system based on the residual network, integrating the Res2Net with the Squeeze-and-Excitation (SE) module. Achieved high accuracy and generalization in detecting synthetic and replay attacks with an EER of less than 1% using the international ASVSpoof dataset.
  • Employed a generative adversarial network to construct a neural vocoder to rebuild voice waveforms, identifying adversarial samples with unknown attack methods, with an AUC of 99% on the VoxCeleb open-source dataset.

Skills

  • Python - Level 3
  • C & C++ - Level 4
  • Git - Level 3
  • Linux - Level 4
  • HTML & CSS - Level 3
  • English - Level 3

Self-evaluation

  • Passionate about technology with a solid professional foundation.
  • Good team collaboration and communication skills.
  • Positive and upward, pursuing excellence, and eager for research and further study.

Trending Tags

usage test