Course Description

It will be a full-day tutorial held on  18 June, 2023.

The planned agenda of the tutorial is as follows:

7:30 – 8:30 am Breakfast

8:30 – 8:45 am

[15 mins] Introduction to the tutorial and the AD problem [Guansong Pang] [slides]

8:45 – 11:45 am

[150 mins] Unsupervised/self-supervised approaches (given only normal training data) 

• 8:45 – 10:15 Video data [Joey Tianyi Zhou, Radu Tudor Ionescu] [slides_Joey, slides_Radu]

10:15 – 10:45 am Coffee break

• 10:45 – 11:45 Image data [Kihyuk Sohn] [slides]

11:45 – 1:00 pm Lunch break

[60 mins] Weakly-supervised AD/ Open-set AD (given some anomaly training examples, in additional to normal training data)

• 1:00 – 2:00 pm Image/video data [Guansong Pang] [slides]

[60 mins] Weakly-supervised approaches  (given video-level labels, to detect frame-level anomalies) 

• 2:00 – 3:00 pm Video data [Yu Tian] [slides]

3:00 – 3:30 pm Coffee break

[60 mins] Anomaly segmentation (with a specific focus on autonomous driving settings)

• 3:30 – 4:30 pm Image data. [Yu Tian] [slides]

[45 mins] Panel on “Anomaly detection: Challenges ahead”. 

• 4:30 – 5:15 pm [All five speakers. Moderator: Radu Tudor Ionescu.]

The recording of the tutorial is now available at YouTube (Apologies that the first talk was not recorded well).

Tutorial Organizers and Speakers

Dr. Guansong Pang, Assistant Professor 

Singapore Management University (SMU), Singapore (gspang@smu.edu.sg) 

Bio: Guansong Pang is a tenure-track Assistant Professor of Computer Science at the School of Computing and Information Systems, Singapore Management University (SMU), Singapore. He obtained his PhD degree at University of Technology Sydney (UTS). His research interest generally lies in machine learning and their applications, with a particular focus on detecting abnormal or unknown instances from different types of data. He has published more than 40 papers in refereed conferences and journals, such as CVPR, ICCV, ECCV, ACM MM, KDD, AAAI, and IJCAI. He is one of the main speakers of KDD’21 and WSDM’21 tutorials on deep anomaly detection, and one of the main organizers of ANDEA and AI4AN workshop series on anomaly and novelty detection at IJCAI and KDD. He also served as a (lead) guest editor of IEEE TNNLS on “deep learning for anomaly detection”. He was named on the prestigious 2020 UTS Chancellor's Award List and the list of The World's Top 2% Scientists (single recent year) released by Stanford University in 2022.

Dr. Joey Tianyi Zhou, Investigator/ Assoc. Prof.(adj.)

A*STAR Centre for Frontier AI Research (CFAR), Singapore (zhouty@cfar.a-star.edu.sg)

Bio: Joey Tianyi Zhou is a senior scientist, Investigator and group manager with A*STAR Centre for Frontier AI Research (CFAR), Singapore.  He is also holding an adjunct faculty position (adj. Assoc. Prof.) at National University of Singapore (NUS). Dr. Zhou received a Ph.D. degree in computer science from Nanyang Technological University (NTU), Singapore. His current interests mainly focus on improving the efficiency and robustness of machine learning algorithms. Dr. Zhou organized ICDCS' 20-21 workshop on Efficient AI meets Edge Computing, ACML'16 workshop on Learning on Big Data workshop and IJCAI'19 workshop on Multi-output Learning; is an Associate Editor for IEEE TETCI and IET Image Processing, and TPC Chair in Mobimedia 2020; and received NeurIPS Best Reviewer Award in 2017.  

Dr. Radu Tudor Ionescu, Professor

University of Bucharest, Romania (raducu.ionescu@gmail.com)

Bio: Radu Tudor Ionescu is a Professor at the University of Bucharest, Romania. He completed his PhD at the University of Bucharest in 2013. He received the 2014 Award for Outstanding Doctoral Research in the field of Computer Science from the Romanian Ad Astra Association. His research interests include machine learning, computer vision, image processing, medical imaging, computational linguistics and text mining. He published over 110 articles at international peer-reviewed conferences and journals, and a research monograph with Springer. He received the "Caianiello Best Young Paper Award" at ICIAP 2013 for the paper entitled "Kernels for Visual Words Histograms". Radu also received the "Young Researchers in Science and Engineering" Prize.

Dr. Yu Tian, Research Fellow

Harvard University (ytian11@meei.harvard.edu)

Bio:  Yu Tian is a postdoctoral research fellow at Harvard University. He received his Ph.D. in computer science at the Australian Institute of Machine Learning (AIML), University of Adelaide. He was also affiliated with South Australian Health and Medical Research Institute (SAHMRI) during his Ph.D. His research interests are in the fields of computer vision and medical image analysis, in particular abnormality and rarity learning, such as image/video anomaly detection for surveillance and industrial applications or early detection of diseases.

Dr. Kihyuk Sohn, Research Scientist

Google Research (kihyuks@google.com) 

Bio: Kihyuk Sohn is a Research Scientist at Google Research. He was a researcher in the Media Analytics group of NEC Laboratories America. He completed his Ph.D. at University of Michigan. He has broad interest in machine learning and computer vision, with particular focus on supervised and unsupervised deep representation learning with applications to computer vision, audio recognition, and text processing, using graphical models that are invariant to many factors of variation for robust perception. 

Contact

Dr. Guansong Pang, gspang@smu.edu.sg