Keynote Speaker Ⅰ

Prof. Hongmin Liu

University of Science and Technology Beijing, China

 

Biography: Hongmin Liu is a professor, doctoral supervisor and Vice Dean at the School of Intelligent Science and Technology, University of Science and Technology Beijing. She is a National Youth Talent. Her main research areas include artificial intelligence, deep learning, and machine vision. She has led two National Key R&D Program (sub-projects) projects, three National Natural Science Foundation projects, two JKW theme projects, and other nine vertical and horizontal projects. The image feature matching and localization technology developed by her team has been applied in Huawei's new-generation "Hetu" application.

She has published 32 academic articles in top journals in the field (20 in the Chinese Academy of Sciences' Zone 1) as the first or corresponding author and two monographs. In the past five years, she has been granted 10 national invention patents as the first inventor (one of which has been transferred), and has registered two software copyrights. She has won the First Prize of the Natural Science Award from the Chinese Association of Automation (ranked 1st), the Second Prize of the Henan Provincial Science and Technology Progress Award (ranked 2nd). She has served as the editorial board member for prestigious international journals such as IEEE T-Cyber, IEEE T-CSVT, IEEE T-FS, and IEEE T-BD. She is also the area chair for the CCF-B level conference ICME-2021/2022/2023, the exhibition chair for PRCV 2022, a member of the Youth Work Committee of the Chinese Association of Automation, and a senior member of the CCF.

Speech Title: TBD

Abstract: TBD

 

Keynote Speaker Ⅱ

Prof. Jie Yang

Shanghai Jiao Tong University, China

 

Biography: Jie Yang received a bachelor’s degree in Automatic Control in Shanghai Jiao Tong University (SJTU), where a master’s degree in Pattern Recognition & Intelligent System was achieved three years later. In 1994, he received Ph.D. at Department of Computer Science, University of Hamburg, Germany. Now he is the Professor and Director of Institute of Image Processing and Pattern recognition in Shanghai Jiao Tong University. He is the principal investigator of more than 30 national and ministry scientific research projects in image processing, pattern recognition, data mining, and artificial intelligence. He has published six books，more than five hundreds of articles in national or international academic journals and conferences. Google citation over 26300，H-index 83. Up to now, he has supervised 5 postdoctoral, 46 doctors and 70 masters, awarded six research achievement prizes from ministry of Education, China and Shanghai municipality.  He has owned 48 patents. Three Ph.D. dissertation he supervised was evaluated as “National Best Ph.D. Dissertation” in 2009, in 2017, in 2019.  He has been chairman and keynote speaker of more than 10 international conferences.

Speech Title: Researches on the Defenses and Out-Of-Distribution Detection in Trustworthy Deep Learning

Abstract: The rapid advancement of deep learning has had a transformative effect on the development of technology and society across a multitude of sectors. In safety-critical contexts, the potential for neural network models to produce unreliable outputs in response to “malicious” or “unanticipated” inputs poses a severe risk. This talk delves into the output reliability from neural network models within the domain of trustworthy deep learning. 1) inputs that involve pixel perturbations, exemplified by adversarial examples,w.r.t the task of adversarial and certified robustness;  2) inputs that represent distribution shifts, exemplified by Out-of-Distribution (OoD) data, w.r.t the task of out-of-distribution detection. We introduce a novel strategy of model augmentation, adopt a multi-head neural network structure, and pose diversity constraints related to adversarial robustness into the model parameters.We adopt a multi-head neural network structure, use the ensemble of multiple heads in place of the ensemble of multiple neural networks,which significantly reduces the computational load in both training and certification phases. We propose that the non-linearity in InD and OoD data hinders PCA from learning a subspace that fully embodies their diversities. We propose a mode ensemble method that not only enhances detection performance but also significantly reduces the performance variance among independent modes.We propose performing linear dimension reduction on the gradient using a designated subspace that comprises principal components. 

 

 

More speakers will be announced soon...