Lei Wang serves as a Research Fellow at the School of Computing, within the ANU College of Engineering, Computing, and Cybernetics. He is under the guidance of Dr. Liang Zheng (ANU) and Prof. Tom Gedeon (Curtin University).
Currently, he also holds the position of Computer Scientist at Active Intelligence Australia Pty Ltd in Perth. In this role, he leads research and development efforts in the field of commercial anomaly detection solutions. His project has successfully raised 1.2 million US dollars, and as a result, he was awarded the Incentive Unit Award.
In addition to his academic and research responsibilities, Lei Wang also contributes to the field as a Visiting Scientist in the Machine Learning Research Group (MLRG) at Data61 (formerly NICTA), a division of The Commonwealth Scientific and Industrial Research Organisation (CSIRO).
Since 2018, Lei has been dedicated to his work as a full-time Computer Vision Researcher with iCetana Pty Ltd in Perth. During this time, he also acted as a Visiting Researcher in MLRG at Data61/CSIRO and contributed as a Visiting Researcher with the Department of Computer Science and Software Engineering at The University of Western Australia (UWA).
Prior to commencing his PhD at ANU and Data61/CSIRO, Lei received several competitive full PhD scholarships, including the ARC Industrial Transformation Training Centre Scholarship, UWA International Fee Scholarship, University Postgraduate Award, Data61 PhD Scholarship, Data61 Top-up Scholarship, and ANU HDR Fee Remission Merit Scholarship.
Throughout his PhD candidature, Lei authored a significant number of first-author papers published in esteemed venues, such as CVPR, ICCV, ECCV, ACM MM, TPAMI, and TIP. His research, conducted under the guidance of Dr. Piotr Koniusz (Data61/CSIRO & ANU), earned him the Sang Uk Lee Best Student Paper Award from ACCV'22.
Lei's research interests encompass a wide range of topics, including action recognition in videos, anomaly detection, video image processing, one- and few-shot learning, deep learning, tensor learning, and domain adaptation.
He is an active member of both IEEE and ACM as a Student Member.
- Doctor of Philosophy - PhD, Computer Science, The Australian National University (22 July 2019 - 12 Dec 2023)
- Master of Professional Engineering - MPE, Software Engineering, The University of Western Australia (29 Feb 2016 - 16 March 2018)
- Active Intelligence Australia PTY Limited, Perth, Western Australia (WA)
- Computer Scientist (contract), 5 July 2021 - present
- Francis and I established the Active Intelligence Research Challenge Award (01/08/2023-present)
- Granted the Incentive Unit Award (09/09/2022-08/09/2032)
- Raised 1.2 million US dollars (~1.79 million Australian dollars)
- Filed a patent [US provisional patent, SN 63/326,525]
- Ebenezer Technologies, Barranquilla, Colombia
- Machine Learning Researcher, 28 Sep 2020 - 28 Jan 2021
- Researched & delivered an anomaly detection system (Alice Guardian smart video surveillance)
- iCetana Pty Ltd, Perth, WA
- Machine Learning / Computer Vision Researcher (Casual Appointment, 20%), R&D, 22 July 2019 - 28 Sep 2020
- Filed a provisional patent [AU Patent AU 2,019,903,775]
- Computer Vision Research Intern (full-time), Research & Development (R&D), 2 July 2018 - 19 July 2019
- Filed a provisional patent [AU Patent AU 2,019,900,316]
- The Australian National University (ANU)
- Research Fellow (full-time), School of Computing, ANU College of Engineering, Computing and Cybernetics, 20 Feb 2023 - present
- The NCI National AI Flagship Merit Allocation Scheme: I am a Lead Chief Investigator (Lead CI) on 'Robust anomaly detection in human-centric videos'
- I am serving as an Area Chair for ACM Multimedia 2024
- Grant: I am a Co-Investigator (Co-I) on 'Review of Xailient’s technical pipeline of facial recognition'
- I am serving as an Area Chair for ICPR 2024
- NCI Adaptater Scheme Q4 2023 (HPC funding scheme): I am a Chief Investigator (CI) on 'Towards building general-purpose multimodal foundation models'
- Grant / Project Award from Oracle (AU$48,000) for Research on 'Automatic, large-scale screening of failure cases in autonomous driving'
- Invited talk on 'Action Recognition: Past, Present and Future' hosted by Harbin Institute of Technology, Shenzhen, China
- Grant: I am a Co-Investigator (Co-I) on 'Sharing early insights for more resilient communities'
- The Commonwealth Scientific and Industrial Research Organisation (CSIRO)
- Visiting Scientist, MLRG, Data61/CSIRO (Black Mountain ACT), 30 June 2023 - present
- Postgraduate Research Student, MLRG, Data61/CSIRO (Acton & Black Mountain ACT), 22 July 2019 - 12 Dec 2023
- Granted the research award of Doctor of Philosophy (01/11/2023)
- Received the Sang Uk Lee Best Student Paper Award from ACCV’22
- One paper each published at CVPR, ICCV, ECCV, ACM-MM, TPAMI & TIP
- Visiting Researcher, Machine Learning Research Group (MLRG), Data61/CSIRO (Floreat WA), 21 Sep 2018 - 19 July 2019
- The University of Western Australia (UWA)
- Visiting Researcher, CSSE/UWA, 10 Oct 2018 - 09 Oct 2019
I would like to express my deepest appreciation to A/Prof. Du Huynh (UWA), Mr. Graeme Woods (now at FinClear), Dr. Moussa Reda Mansour (now at Alcoa), Dr. Piotr Koniusz (Data61/CSIRO & ANU) and Dr. Liang Zheng (ANU).
- Action recognition
- Anomaly detection
- Video image processing
- One- and few-shot learning
- Deep learning
- Tensor learning
- Domain adaptation
|3Mformer: Multi-order Multi-mode Transformer for Skeletal Action Recognition
L Wang, P Koniusz
Computer Vision and Pattern Recognition (CVPR), 5620-5631
|Temporal-Viewpoint Transportation Plan for Skeletal Few-shot Action Recognition
L Wang, P Koniusz
Asian Conference on Computer Vision (ACCV), 307-326
|[B, oral, 4.9% acceptance rate,
Best Student Paper Award]
|Uncertainty-DTW for Time Series and Sequences
L Wang, P Koniusz
European Conference on Computer Vision (ECCV), 176-195
|[A*, oral, 2.7% acceptance rate]
|Fusing Higher-Order Features in Graph Neural Networks for Skeleton-Based Action Recognition
Z Qin, Y Liu, P Ji, D Kim, L Wang, B McKay, S Anwar, T Gedeon
IEEE Transactions on Neural Networks and Learning Systems (TNNLS), (Early Access)
|Self-supervising Action Recognition by Statistical Moment and Subspace Descriptors
L Wang, P Koniusz
ACM International Conference on Multimedia (ACM-MM), 4324-4333
|Tensor Representations for Action Recognition
P Koniusz, L Wang, A Cherian
IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI) 44 (2), 648-665
|A Comparative Review of Recent Kinect-based Action Recognition Algorithms
L Wang, DQ Huynh, P Koniusz
IEEE Transactions on Image Processing (TIP) 29 (1), 15-28
Dataset & Code
|Hallucinating IDT Descriptors and I3D Optical Flow Features for Action Recognition with CNNs
L Wang, P Koniusz, DQ Huynh
IEEE International Conference on Computer Vision (ICCV), 8698-8708
|Loss Switching Fusion with Similarity Search for Video Classification
L Wang, DQ Huynh, MR Mansour
26th IEEE International Conference on Image Processing (ICIP), 974-978
|[B, industrial research
+ 1 AU patent]
|System and Method of Detecting Anomalies from Mass Data
US patent (provisional, SN 63/326,525)
|Method and System for Classifying Video Data
L Wang, G Woods
AU Patent AU 2,019,903,775
|System and Method of Video Data Retrieval
L Wang, M Reda Mansour, G Woods
AU Patent AU 2,019,900,316
|Robust Human Action Modelling
PhD thesis*, The Australian National University
|Analysis and Evaluation of Kinect-based Action Recognition Algorithms
Master’s thesis, The University of Western Australia
*: I would like to extend my heartfelt gratitude to three esteemed PhD thesis examiners, namely Dr. Lingqiao Liu (University of Adelaide) and Dr. Peng Wang (University of Electronic Science and Technology of China), as well as an anonymous examiner, for their invaluable insights, meticulous examination, and constructive feedback on my research work.
|Flow Dynamics Correction for Action Recognition
L Wang, P Koniusz
arXiv preprint arXiv:2310.10059
|Adaptive Multi-head Contrastive Learning
L Wang, P Koniusz, T Gedeon, L Zheng
arXiv preprint arXiv:2310.05615
|3D Skeleton-based Few-shot Action Recognition with JEANIE is not so Naïve
L Wang, J Liu, P Koniusz
arXiv preprint arXiv:2112.12668
|(an extended version has been
accepted by ACCV'22 [oral])
|High-order Tensor Pooling with Attention for Action Recognition
L Wang, P Koniusz, K Sun
arXiv preprint arXiv:2110.05216
|(an extension of our TPAMI'20)
|AI in Software Engineering: Case Studies and Prospects
arXiv preprint arXiv:2309.15768
|Technical Report. The author conducted this work while enrolled as a master's student at UWA, specifically for the CITS5502 Software Processes unit in 2017.||2017|
Light cyan highlights the research works during my PhD candidature and light yellow highlights my industrial research works. My theses are in misty rose color and other collaborative research works are in platinum color.
I am serving as an Area Chair for ACM Multimedia 2024.
I have the honor of serving as an Area Chair for the 27th International Conference on Pattern Recognition (ICPR 2024) by invitation from Prof. Cheng-Lin Liu (CASIA).
I regularly review technical papers for the following journals, conferences & workshops.
- IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), International Journal of Computer Vision (IJCV),
- IEEE Transactions on Circuits and Systems for Video Technology (TCSVT), IEEE Transactions on Multimedia (T-MM), Pattern Recognition (PR), Computer Vision and Image Understanding (CVIU),
- Neural Computing and Applications (NCAA), Journal of Visual Communication and Image Representation (JVCI), Signal, Image and Video Processing (SIVP),
- The Visual Computer (TVCJ), IET Computer Vision, Electronics Letters
- ICLR 2024, AAAI 2024, NeurIPS 2023, ICML 2023, ICLR 2023, AAAI 2023, NeurIPS 2022, AAAI 2022, ICML 2022,
- ICCV 2023, CVPR 2023, CVPR 2022, ECCV 2022 (outstanding reviewer),
- WACV 2024, BMVC 2023, BMVC 2022, BMVC 2021, BMVC 2020,
- IEEE Conference on Artificial Intelligence (IEEE CAI 2023)
- Deep Vision Workshop (CVPR 2020),
- Vision Datasets Understanding (1st VDU at CVPR 2022, 2nd VDU at CVPR 2023),
- The AI City Challenge Workshop at CVPR 2023
I also serve as a thesis examiner for master's and honors theses at the ANU College of Engineering, Computing & Cybernetics for the academic terms S1 2023 and S2 2023.
- The University of Western Australia (UWA)
|Action Recognition: Past, Present and Future [slides]||Prof. Haijun Zhang,
Department of Computer Science,
Harbin Institute of Technology,
|"The next generation of international Chinese young students face to face" Issue 21
International Cooperation and Exchange Program Series Activities of Harbin Institute of Technology (Shenzhen)
|12 Aug 2023|
|Modeling Videos: Language as a Key Driver [slides]||Dr. Liang Zheng's weekly paper reading session||This seminar explores cutting-edge video-language models and their applications in downstream video processing tasks, including action recognition, action localization, and text-video retrieval. It spans closed-set, few-shot, and zero-shot scenarios.||6 Dec 2023|
|Industry-Academia Dynamics in Computer Vision: Trends, Collaboration, and Citation Patterns [slides]||Dr. Liang Zheng's weekly paper reading session||This reading session covers (i) research trend in computer science, (ii) industry & academic research in computer vision and (iii) a closer look at citations.||8 Sep 2023|
|Contrastive Learning meets Masked Modeling [slides]|| Dr. Liang Zheng's weekly paper reading session
 Dr. Piotr Koniusz's research group (repeated paper reading session)
|This reading session covers a comparison between contrastive learning and masked image modeling with regard to model behavior, model architecture, self-attention, and representation.||2 June 2023|
|Robust Human Action Modelling [slides]||ANU College of Engineering, Computing and Cybernetics||PhD completion seminar: This seminar covers the following topics (i) an introduction to action recognition (AR) and a comparative review of AR methods, (ii) video-based action recognition, (iii) skeleton-based action recognition, and (iv) one- and few-shot action recognition.||2 Feb 2023|
|Analysis and Evaluation of Kinect-based Action Recognition Algorithms [slides]||UWA's Department of Computer Science and Software Engineering||MPE Engineering Research Project: In this seminar, I discussed (i) applications, issues, and techniques in action recognition (AR) and (ii) an analysis and evaluation of four handcrafted AR algorithms.||Oct 2017|