Jun Jie (Jackson) Ouyang
Computer Scientist & Basketball Enthusiast
About
Hello! I'm Jackson, a passionate Computer Vision and Machine Learning researcher currently pursuing my M.S. in Computer Science at USC. I'll be starting my Ph.D. at Dartmouth College in September 2025.
My research spans multiple exciting frontiers in AI: document understanding, video understanding, generative AI, and computer graphics. I'm particularly fascinated by multimodal transformers, video-language models, neural rendering, and the intersection of computer vision with generative systems. My work in document AI has taught me the power of multimodal fusion, which I'm now extending to video understanding and 3D graphics generation.
Beyond traditional research, I'm driven by a bold vision: creating an intelligent agent system inspired by TARS and CASE from Interstellar. I envision building an AI companion that combines advanced reasoning, multimodal perception, natural conversation, and adaptive personality—a system that can truly understand and interact with the world in meaningful ways. This ambitious goal motivates my exploration of generative AI, embodied intelligence, and human-AI interaction.
When I'm not deep in research or coding, you'll find me on the basketball court. Sports, especially basketball, help me stay balanced and bring a competitive edge to everything I do. I believe the teamwork and strategic thinking from sports translate beautifully into collaborative research and building intelligent systems.
I'm passionate about pushing the boundaries of what AI can achieve—from understanding complex video narratives to generating photorealistic graphics, from parsing documents to creating genuinely helpful AI companions. I've had the privilege of contributing to multiple top-tier conferences including CVPR 2025 and ICPR 2024, and I'm excited to continue exploring the future of AI.
Publications
Runbo Zhao, Junjie Ouyang, Chen Gao, Xugong Qin, Gangyan Zeng, Xiaoxu Hu, and Peng Zhang, "Perception‑Enhanced Generative Transformer for Key Information Extraction from Documents," Springer, Cham, 2024.
Research Experience
Research Assistant with Assistant Professor Yue Zhao
University of Southern California • Sep 2023 – Dec 2024
- Brainstormed with peers to improve PyGOD, an open-source Python library for graph outlier detection
- Assisted in presenting theoretical models proposed in the research group
- Modified Python scripts and wrote new modules for PyGOD; conducted feasibility and accuracy tests
- Drafted documentation and reports summarizing findings and proposed changes
Research Assistant with Professor Peng Zhang
Institute of Computing Technology, Chinese Academy of Sciences • Apr 2023 – Present
- Developed Word Detection Algorithms: Designed modules for structured text in complex layouts under PEGT project
- Enhanced Multimodal Data Processing: Integrated visual and textual data using pre-trained models
- Evaluated PEGT: Benchmarked on SROIE and CORD; showed F-measure gains; published at ICPR 2024
- Published on PaddlePaddle: Sole author of PaddleDetection module paper (IIoTBDSC 2024)
- Optimized YOLO and Faster R-CNN models with data augmentation for robustness
- Co-authored CVPR 2025 papers on CLIP and document retrieval
Research Assistant with Professor Xin Rong Chen
West China Medical School, West China Hospital, Sichuan University • Dec 2023 – Jun 2024
- Researched VR for postoperative pain management; explored patient distraction and anxiety reduction
- Contributed introduction on VR technology for publication in Journal of Mechanics in Medicine and Biology
- Reviewed combined therapies integrating VR with traditional approaches
Research Assistant with Professor Tae Oum
University of British Columbia • Sep 2022 – Dec 2022
- Analyzed impacts of air transport connectivity in China; compiled and processed large SQL datasets
- Conducted econometric analysis using Lasso regression
Research Assistant with Professor Jinyuan Zhang
University of California, Los Angeles • Apr 2022 – Aug 2022
- Conducted industry research on US fintech companies
- Scraped and preprocessed data with Python and SQL
- Compared business models and evaluated market impact
Undergraduate Researcher with Professor Mark Vogelsberger
Massachusetts Institute of Technology • Jun 2021 – Aug 2021
- Built an NLP platform for sentiment analysis on social media reactions
- Managed databases using SQL and MongoDB
- Co-authored sentiment analysis paper submitted to CSTMM 2022
Undergraduate Researcher
Huazhong University of Science and Technology • Jan 2021 – May 2021
- Assessed marketing strategies and consumer behavior
- Built a web interface (HTML/JS) for presenting statistical results
- Designed logistic regression models in Python; increased product sales by 20%
Data Analyst Intern
MOYI Tech • Jul 2022 – Sep 2022
- Evaluated the relationship between social media sentiment and market trends
- Scraped 10,000+ comments with Python; built pipeline to predict index returns
- Improved prediction accuracy by 50%