Prof. Daeseon Choi(SonngsilUniversity)

Title : AI and Privacy

Abstract :
These days, many AI technologies are used in various environments and applications. Some AI models are trained with private data. And some AI systems are used for gathering private information, making an inference of private attributes, and tracking user's activities.

Therefore AI has many privacy issues. In this talk, I will introduce four kinds of privacy issues of AI. They are

1) training data privacy 2) private information leakage from a deployed AI model 3) AI devices that gather private information 4) AI technologies that threaten privacy.

For each issue, I will explain what the problem is and how we mitigate such a problem.

Invited Talks

Invited Talks #1 (10:30~11:15, Aug. 19)

Title: IoT Cyber Security: Device Discovery, Attack Observation, and Security Notification


  • Prof. Katsunari Yoshioka (Yokohama National University)

    Short Bio:
    Katsunari Yoshioka is an Associate Professor at Yokohama National University since 2011. Before that, he was a researcher at National Institute of Information and Communications Technology, Japan. His research interests covers wide area of system security and network security including malware analysis and IoT security. He received the commendation for science and technology by the minister of MEXT, Japan in 2009, the award for contribution to Industry-Academia-Government Collaboration by the minister of MIC, Japan in 2016, and the Culture of Information Security Award in 2017.

    Abstract :
    IoT cyber security has become one of the most important and challenging topics in recent years. In this talk, new trends in IoT cyber attacks, malware evolution, and efforts to discover and mitigate insecure and/or compromised devices are explained.

  • Invited Talk #2 (15:15~16:00, Aug. 19)

    Title: Attacks against Transfer Learning and their Countermeasures


  • Prof. chia-Mu Yu
  • Department of Information Management and Finance
  • National Yang Ming Chiao Tung University
  • Taiwan

    Abstract :
    Transfer learning is a powerful approach that allows users to quickly build accurate deep-learning models by "learning" from centralized models pre-trained with large datasets. However, the centralization of model training increases their vulnerability to misclassification attacks leveraging the knowledge of publicly accessible pre-trained models. This talk will have a brief overview of the techniques that allow attackers to launch attacks, such as misclassification and brute force. We will also describe the candidate mitigations to the attacks.

  • Invited Talk #3 (10:00~10:45, Aug. 20)

    Title: Data driven analysis for cybersecurity


  • Dr. Daisuke Inoue
  • Director General of Cybersecurity Nexus
  • National Institute of Information and Communications Technology

    Short Bio :
    Daisuke Inoue received the B.E. and M.E. degrees in electrical and computer engineering and the Ph.D. degree in engineering from Yokohama National University, Yokohama, Japan in 1998, 2000, and 2003, respectively. He joined Communications Research Laboratory (CRL), Japan, in 2003. CRL was relaunched as National Institute of Information and Communications Technology (NICT) in 2004, where he is currently both the Director General of Cybersecurity Nexus and the Director of Cybersecurity Laboratory. He received several awards including the best paper award at the 2002 Symposium on Cryptography and Information Security (SCIS 2002), the commendation for science and technology by the Minister of MEXT, Japan, in 2009, the Good Design Award 2013, the Asia-Pacific Information Security Leadership Achievements (ISLA) 2014, the award for contribution to Industry-Academia-Government Collaboration by the Minister of MIC, Japan, in 2016, the Maejima Hisoka Award, in 2018, and NDSS 2019 Distinguished Paper Award at the Network and Distributed System Security Symposium (NDSS) 2019.

    Invited Talk #4 (15:15~16:00, Aug. 20)

    Title: Security and Privacy for Artificial Intelligence of Things

    Speaker: Prof. Haojin Zhu (Shanghai Jiao Tong University)


    Short Bio:
    Haojin Zhu received the Ph.D. in Electrical and Computer Engineering from the University of Waterloo (Canada), in 2009. He is currently a professor with Computer Science & Engineering department in Shanghai Jiao Tong University. His research interests include network security and privacy-enhancing technologies. He published more than 60 international journal papers, including JSAC, TDSC, TPDS, TMC, TIFS, and 80 international conference papers, including IEEE S&P, ACM CCS, USENIX Security, NDSS, ACM MOBICOM, ACM MOBIHOC. He received a number of awards including IEEE TCSC Award for Excellence in Scalable Computing (Middle Career Researcher, 2020),Natural Science Award of Ministry of Education (2018),IEEE ComSoc Asia-Pacific Outstanding Young Researcher Award (2014), Top 100 Most Cited Chinese Papers Published in International Journals (2014), Supervisor of Shanghai Excellent Master Thesis Award(2014), Distinguished Member of the IEEE INFOCOM Technical Program Committee (2015, 2020). He was a co-recipient of best paper awards of IEEE ICC (2007) and Chinacom (2008), IEEE GLOBECOM Best Paper Nomination (2014), WASA Best Paper Runner-up Award (2017). He received Young Scholar Award of Changjiang Scholar Program by Ministry of Education of P.R. China in 2016. He is serving on the associate editorial board for IEEE Trans. on Wireless Communications and program committees for ACM CCS, USENIX Security, NDSS, IEEE INFOCOM.

    Abstract :
    Artificial Intelligence of Things (AIoT), a novel technology of combining Artificial intelligence (AI) technologies with the Internet of things (IoT), is transforming the way we interact with our devices at home, at work, and throughout our cities to achieve more efficient IoT operations. However, AIoT is also opening a door for the adversary to launch the attacks by exploiting the vulnerabilities of the platform, human-computer interfaces or AI control algorithms. In this study, we will introduce our latest research progress on AIoT security and privacy including state-of-the-art, attacks, and defenses.

    Invited Talk #5 (17:40~18:25, Aug. 20)

    Title: Computation and Communication Security for the Fifth Generation and Beyond (B5G) Mobile Networks


  • Prof. Ruei-Hau Hsu
  • Department of Computer Science and Engineering
  • National Sun Yat-sen University
  • Taiwan

    Abstract :
    The fifth generation (5G) mobile networks enable new features of communications and computation models, i.e., device-to-device (D2D) communications, massive machine-type communications (mMTC), and multi-access edge computation (MEC), etc. By the above new features, B5G mobile networks enable IoT applications for higher bandwidth utilization, ultra-reliable and low latency communications, and machine-type communications for a massive number of devices. However, the new types of communications and computation models become more localized and distributed with a considerable number of devices. In such a scenario, it is adverse to the deployment of security mechanisms, such as authentication for access control of services, and session key exchange for confidentiality and integrity of communications, where centralized trusted authorities or authentication servers are usually adopted. The traditional key management, which is centralized, will increase the latency and reduce the availability in nature. Moreover, the system model of new types of IoT services based on B5G mobile networks are more comprehensive, where users of IoT services, service providers, and mobile network operators are not fully trusted to each other. Thus, privacy protection is more critical to the construction of security mechanisms for D2D communications, mMTC, and MEC. In this talk, we will present the design of new security systems in B5G by certain enhanced cryptosystems, such as attribute-based encryption, computation-outsourced cryptosystem, encryption controllable cryptosystem, etc., for fine-grained access control and authorization with privacy protection. Overall, the studies pave the ways for the development of the provisioning systems in communications and computation for IoT applications in B5G mobile networks.

  • Contact us: ⓒ AsiaJCIS 2021