About Me

Hi, I’m Zhenxiang (Roy) Jiang, an Applied Scientist/AI Engineer/Researcher.

With over three years of experience in deep learning R&D and deployment, my work spans a wide range of artificial intelligence tasks, from computer vision to large language model.

Portfolio

CV/Resume

Research & Development Areas

• Video Agent
• Image/Video Understanding
• Large language model
• Image generation & editing
• Detection and classification
• Human-related tasks
• 3D/4D dynamic scene reconstruction
• Explainable/trustworthy AI

Professional Experience

OpusClip

AI Engineer/Applied Scientist
Direct Manager: Vito Zhu
Palo Alto, California, US | May 2025 - Present

• Contributing to the development of Agent Opus, which is an AI video agent that turns your ideas in any form into polished videos.
• Focusing on integrating real-world assets, including images, webpages, videos, and posts, into video generation pipelines.
• Led the design and implementation of core workflows for agent action and evaluation, with a particular emphasis on image and video understanding and their evaluation across different agents.
• Designed workflows for atomic template creation, seamlessly integrating real-world assets with image/video generation and editing capabilities.
• Researched the ability of integrating the ability of detection, segmentation with LLM-based agent system.

Learning and Vision Lab, ECE Dept., National University of Singapore

Research Assistant
Supervisor: Prof. Xinchao Wang
Singapore | August 2023 – February 2025

• Completed a diverse range of computer vision tasks, from low-level image processing to high-level scene understanding.
• Co-led a high-resolution non-homogeneous dehazing project that ranked 4th out of 100+ submissions (CVPR Workshop 2023).
• Collaborated on an XAI project with Singapore’s largest national defense R&D organization, delivering two phases of product development.
• Designed key modules—camera–world coordinate conversion and interactive 3D/4D visualization—for the GFlow and C4D projects, contributing to publications at AAAI 2025 and arXiv.

Temasek Laboratories, National University of Singapore

Research Assistant
Supervisor: Dr. Sunan Huang
Singapore | September 2023 – April 2024

• Led research on a high-frequency drone detection module to enhance onboard drone tracking system accuracy.
• Built a fully labeled event camera drone detection dataset by integrating multiple drone detection datasets.

Machine Intelligence Lab, College of Computer Science, Sichuan University

Research Assistant
Supervisor: Prof. Yuanyuan Chen
Chengdu, China | March 2022 – June 2023

• Initiated research on facial expression recognition under face mask occlusion and developed a seven-class dataset, earning the Best Presentation Award at ACM ICCAI 2023.
• Led the development of WS-GCN for weakly supervised 3D human pose estimation, resulting in a publication at ACM ICCAI 2024.

Yinlaiyinwang (Convenient Printing) Technology

Founder, CEO
Chengdu, China | November 2020 – July 2022

• Led a team of 10 to develop an intelligent online printing system, converting traditional offline printers into smart, internet-connected devices.
• Established an on-campus experience store serving over 100,000 students and creating more than 15 part-time job opportunities.
• Received multiple entrepreneurship awards at the college and university levels.

Education

National University of Singapore

Master of Science in Computer Engineering
Specialization: Machine Intelligence and Application
GPA: 4.69/5.00
Singapore | August 2023 – January 2025

Sichuan University

Bachelor of Engineering in Artificial Intelligence
GPA: 3.80/4.00 | Top Graduate of Sichuan Province (Top 4%) | Graduated as Valedictorian
Chengdu, Sichuan, China | September 2019 – June 2023

Publications

• C4D: 4D Made from 3D through Dual Correspondences
  Wang, S., Jiang, Z., Yang, X., & Wang, X.
  ICCV 2025 (Accepted)
  Paper

• GFlow: Recovering 4D World from Monocular Video
  Wang, S., Yang, X., Shen, Q., Jiang, Z., & Wang, X.
  In Proceedings of the AAAI Conference on Artificial Intelligence (Vol. 39, No. 8, pp. 7862-7870), April 2025
  Paper

• WS-GCN: Integrating GCN with Weak Supervision for Enhanced 3D Human Pose Estimation
  Jiang, Z., Chen, Y.
  In Proceedings of the 2024 10th International Conference on Computing and Artificial Intelligence (pp. 6–13), April 2024
  ACM Digital Library

• NTIRE 2023 HR Nonhomogeneous Dehazing Challenge Report
  Ancuti, C. O., …, Wu, Y., Jiang, Z., Liu, S., Yang, X., Jing, Y., … & Busch, C.
  In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (pp. 1808–1825), 2023
  Paper

• A Novel Seven-Class Facial Expression Recognition Method With Face Mask
  Jiang, Z.
  In Proceedings of the 2023 9th International Conference on Computing and Artificial Intelligence (pp. 178–184), March 2023
  ACM Digital Library

Skills

Programming Languages

• Python (Advanced)
• SQL (Advanced)
• C++ (Proficient)
• Matlab (Proficient)
• CudaC (Proficient)
• Java (Intermediate)
• Shell Scripting (Intermediate)

Libraries & Frameworks

• PyTorch (Advanced)
• NumPy (Advanced)
• Pandas (Advanced)
• Matplotlib (Proficient)
• Scikit-Learn (Proficient)
• OpenCV (Proficient)
• TensorBoard (Familiar)
• LaTeX (Familiar)

Tools & Platforms

• Linux (Advanced)
• MySQL (Advanced)
• Git (Advanced)
• Docker (Proficient)
• FastAPI (Proficient)
• Nginx (Familiar)
• Vue (Familiar)
• GitHub Actions (Familiar)

Languages

• English (Fluent)
• Mandarin Chinese (Native)