profile picture

Dehai Zhao

赵德海

Education

PhD in Computer Science

Australian National University
2018 - 2022

M.Eng in Software Engineering

China University of Petroleum (East China)
2014 - 2017

B.Eng in Software Engineering

China University of Petroleum (East China)
2010 - 2014

Profile

I am a researcher at Zhejiang University, where I focus on Large Language Models, Embodied Intelligence, and AI Agents. Previously, I worked as a Research Scientist at CSIRO’s Data61 in the SE4AI team, where I led the development of guardrail systems for Large Language Models, a key component of Responsible AI.

I obtained my PhD from the Research School of Computer Science at the Australian National University, with a primary focus on intelligent software engineering. My doctoral research explored harnessing computer vision and deep learning techniques to address software engineering challenges, aiming to elevate the quality and efficiency of programming. I also maintain a keen interest in cloud computing and the Internet of Things (IoT).

Projects

My projects mainly focus on two areas: programming screencast analysis and automatic GUI testing. All of them are based on computer vision and deep learning methods.

SeeHow - Using computer vision and deep learning methods to extract line granularity coding behaviours from programming screencasts.
Seenomaly - Computer vision and deep learning based methods to solve GUI animation linting problem.
ActionNet - Applying computer vision and deep learning methods to extract primary actions such as moving mouse and typing text from programming screencasts.

Publications

  • Hard to Read and Understand Pythonic Idioms? DeIdiom and Explain Them in Non-Idiomatic Equivalent Code
    • Zejun Zhang, Zhenchang Xing, Dehai Zhao, Qinghua Lu, Xiwei Xu, Liming Zhu
    ICSE2024 (ACM SIGSOFT Distinguished Paper Award)
  • SeeHow: Workflow Extraction from Programming Screencasts through Action-Aware Video Analytics.
    • Dehai Zhao, Zhenchang Xing, Deheng Ye, Xiwei Xu, Liming Zhu
    ICSE2023
  • Distinguishing Look-Alike Innocent and Vulnerable Code by Subtle Semantic Representation Learning and Explanation.
    • Chao Ni, Xin Yin, Kaiwen Yang, Dehai Zhao, Zhenchang Xing, Xin Xia
    FSE2023
  • From Misuse to Mastery: Enhancing Code Generation with Knowledge-Driven AI Chaining.
    • Xiaoxue Ren, Xinyuan Ye, Dehai Zhao, Zhenchang Xing, Xiaohu Yang
    ASE2023
  • A Streaming Cloud Platform for Real-Time Video Processing on Embedded Devices.
    • Weishan Zhang, Haoyun Sun, Dehai Zhao, Liang Xu, Xin Liu, Huansheng Ning, Jiehan Zhou, Yi Guo, Su Yang
    IEEE Transactions on Cloud Computing 9(3): 868-880 (2021)
  • CNN-based Multi-model Birdcall Identification on Embedded Devices.
    • Shidong Pan, Dehai Zhao, Weishan Zhang
    SmartIoT 2021: 245-251
  • Seenomaly: Vision-Based Linting of GUI Animation Effects Against Design-Don’t Guidelines.
    • Dehai Zhao, Zhenchang Xing, Chunyang Chen, Xiwei Xu, Liming Zhu, Guoqiang Li, Jinshui Wang
    ICSE 2020
  • ActionNet: vision-based workflow action recognition from programming screencasts.
    • Dehai Zhao, Zhenchang Xing, Chunyang Chen, Xin Xia, Guoqiang Li
    ICSE 2019: 350-361 (ACM SIGSOFT Distinguished Paper Nomination)
  • Multi-source data fusion using deep learning for smart refrigerators.
    • Weishan Zhang, Yuanjie Zhang, Jia Zhai, Dehai Zhao, Liang Xu, Jiehan Zhou, Zhongwei Li, Su Yang
    Computers in Industry 95: 15-21 (2018)
  • An intelligent power distribution service architecture using cloud computing and deep learning techniques.
    • Weishan Zhang, Gaowa Wulan, Jia Zhai, Liang Xu, Dehai Zhao, Xin Liu, Su Yang, Jiehan Zhou
    J. Network and Computer Applications 103: 239-248 (2018)
  • Emotion Recognition from Chinese Speech for Smart Affective Services Using a Combination of SVM and DBN.
    • Lianzhang Zhu, Leiming Chen, Dehai Zhao, Jiehan Zhou, Weishan Zhang
    Sensors 17(7): 1694 (2017)
  • Deep learning and SVM-based emotion recognition from Chinese speech for smart affective services.
    • Weishan Zhang, Dehai Zhao, Zhi Chai, Laurence T. Yang, Xin Liu, Faming Gong, Su Yang
    Softw., Pract. Exper. 47(8): 1127-1138 (2017)
  • Patching by automatically tending to hub nodes based on social trust.
    • Xin Liu, Yao Wang, Dehai Zhao, Weishan Zhang, Leyi Shi
    Computer Standards & Interfaces 44: 94-101 (2016)
  • Deep Learning Based Emotion Recognition from Chinese Speech.
    • Weishan Zhang, Dehai Zhao, Xiufeng Chen, Yuanjie Zhang
    ICOST 2016: 49-58
  • Workload Prediction for Cloud Cluster Using a Recurrent Neural Network.
    • Weishan Zhang, Bo Li, Dehai Zhao, Faming Gong, Qinghua Lu
    IIKI 2016: 104-109
  • An Empirical Study on Big Video Data Processing: Architectural Styles, Issues, and Challenges.
    • Weishan Zhang, Zhichao Wang, Liang Xu, Dehai Zhao, Faming Gong, Qinghua Lu
    IIKI 2016: 110-115
  • Distributed embedded deep learning based real-time video processing.
    • Weishan Zhang, Dehai Zhao, Liang Xu, Zhongwei Li, Wenjuan Gong, Jiehan Zhou
    SMC 2016: 1945-1950
  • A Distributed Video Management Cloud Platform Using Hadoop.
    • Xin Liu, Dehai Zhao, Liang Xu, Weishan Zhang, Jijun Yin, Xiufeng Chen
    IEEE Access 3: 2637-2643 (2015)
  • Food Image Recognition with Convolutional Neural Networks.
    • Weishan Zhang, Dehai Zhao, Wenjuan Gong, Zhongwei Li, Qinghua Lu, Su Yang
    UIC/ATC/ScalCom 2015: 690-693

    Awards

    ICSE 2025 ACM SIGSOFT Distinguished Paper Award

    ICSE 2024 ACM SIGSOFT Distinguished Paper Award

    ICSE 2019 Distinguished Paper Nomination Award

    2017 Honors Graduate of China University of Petroleum

    2016 CCF Big Data & Computing Intelligence Contest Special Award.

    Services

    FSE 2025 Program Committee

    ASE 2025 Program Committee

    ASE 2024 Program Committee

    MSR 2021 Shadow Program Committee

    Teaching

    2020 Semester 1 | Tutor | Software Engineering

    2020 Semester 2 | Tutor | Software Engineering

    2021 Semester 1 | Lecture | Software Engineering