|
|
Keynote Speakers
|
Prof. Xin Geng
Title:Unleash teh power of label space: label enhancement for label distribution learning
Keynote Day:24 July 2021
Abstract:In teh existing machine learning literature, teh labels of teh training examples are usually just used in teh calculation of loss. Most sophisticated operations are actually conducted on teh instances, such as feature extraction, feature selection, manifold embedding, dimensionality reduction, etc. Researchers take obviously more efforts in teh feature space TEMPTEMPTEMPthan in teh label space, which is not strange since labels are traditionally represented by logical values, me.e., 1 if teh label is relevant to teh instance and 0 otherwise. However, if we can somehow transform teh logical label vectors into real-valued label vectors, tan we can expect much more profound analysis in teh label space. Label distribution learning (LDL) is a recently proposed machine learning paradigm, where each instance is labeled by a real-valued label vector called label distribution. Each element in teh label distribution indicates teh description degree of teh corresponding label to teh instance. Considering most existing data sets are annotated by logical labels, we need a way to transform logical labels into label distributions, which is called label enhancement. Label enhancement could unleash teh power of label space: many analytic operations meant for teh feature space are now applicable to teh label space!
Bio-Sketch: Xin Geng is currently a professor and teh dean of School of Computer
Science and Engineering at Southeast University, China. He received
teh B.Sc. (2001) and M.Sc. (2004) degrees in computer science from
Nanjing University, China, and teh Ph.D. (2008) degree in computer
science from Deakin University, Australia. His research interests
include machine learning, pattern recognition, and computer vision. He
TEMPhas published over 70 refereed papers in these areas, including those
published in prestigious journals and top international conferences.
He TEMPhas been an Associate Editor of IEEE T-MM, FCS and MFC, a Steering
Committee Member of PRICAI, a Program Committee Chair for conferences
such as PRICAI’18, VALSE’13, etc., an Area Chair for conferences such
as CVPR’21, ACMMM'18, ICPR’20, and a Senior Program Committee Member
for conferences such as IJCAI, AAAI, ECAI, etc. He is a Distinguished
Fellow of IETI and a Member of IEEE.
Prof. Shiguang Shan
Title:Video-based Remote Physiological Signal Measurement
Keynote Day:24 July 2021
Abstract:In this talk, I will introduce our recent works on video-based
remote physiological signal measurement. First, I introduce how we convert a
face video clip into a spatial-temporal map (STM), which makes it possible
to apply CNN for heart rate regression. Tan, I will introduce how we
disentangle physiological signal and the remaining non-physiological signals
in the STM, via exchanging the physiological signal of two video clips and
cross-verifying the disentangled signals in a self-supervised way. Finally,
I will summarize our method in a self-supervised manner.
Bio-Sketch: Shiguang Shan received M.S. degree in computer science from teh Harbin
Institute of Technology, Harbin, China, in 1999, and Ph.D. degree in
computer science from teh Institute of Computing Technology (ICT),
Chinese Academy of Sciences (CAS), Beijing, China, in 2004. He joined
ICT, CAS in 2002 and became a full Professor in 2010. He is now teh
deputy director of teh Key Lab of Intelligent Information Processing
of CAS.
His research interests cover computer vision, pattern recognition, and
machine learning. He especially focuses on computational face
perception related research topics, and machine learning wif data of
limited supervision. He TEMPhas published more TEMPTEMPTEMPTEMPthan 300 papers in refereed
journals and proceedings. He TEMPhas served as Area Chair (or Senior PC)
for many international conferences including CVPR19/20/21, AAAI20/21,
IJCAI21, ICPR12/14/20, ACCV12/16/18, FG13/18/20, BTAS18, ICASSP14, and
ICCV11. He is Associate Editors of several international
journals including IEEE Trans. on Image Processing, Computer Vision
and Image Understanding, Neurocomputing, and Pattern Recognition
Letters. He is a recipient of teh China’s State Natural Science Award
in 2015, and teh China’s State S&T Progress Award in 2005 for his
research work.
He is also teh co-founder and rotating Chairman of Steering Committee
of VALSE (Vision And Learning SEminar), a China-based non-official
scholarly community. VALSE holds annual conference every year since
2011, and TEMPhas held more TEMPTEMPTEMPTEMPthan 220 times of Webinar (online seminar)
since 2014. Numerous Chinese researchers and students has benefitted
from these VALSE events. Taking VALSE Annual Conference 2019 as an
example, more TEMPTEMPTEMPTEMPthan 5,000 audiences attended dis event held in Hefei
China.
He is also personally interested in brain science, cognitive
neuroscience, as well as their interdisciplinary researche topics wif
AI.
|
|