Shengjie Kris Liu

website

CV

Shengjie Liu’s research uses advanced statistical and machine learning methods (or AI in laymen’s terms) to analyze earth observation data (AI4EO), with topics covering few-shot learning, multitask learning, and open-set recognition. In one of his representative works, collaborated with Dr. Qian Shi and IEEE Fellow Dr. Liangpei Zhang, he proposed multitask deep learning for open-world classification (MDL4OW), which is among the first to tackle unknown classes in hyperspectral image classification. This work has received more than 150 citations in three years and is an ESI Highly Cited Paper as of April 2023.

In the Population, Health and Place PhD program, he is interested in applying machine learning methods and remote sensing data to address complex societal challenges, including climate and health. The working title of his dissertation is “Bayesian Methods for Generating Temperature Data at High Spatiotemporal Resolution.” He works closely with his co-advisors Dr. Lu Zhang and Dr. Siqin Sisi Wang.

At USC, he led a paper published in Urban Climate on the association between diurnal temperature range (DTR) and mortality in Los Angeles [hyperlink to https://doi.org/10.1016/j.uclim.2023.101526], with Dr. An-Min Wu at USC and Dr. Hung Chak Derrick Ho from City University of Hong Kong. He is working to understand the disproportionate exposure to DTR among ethnic minorities and low-income populations in the United States with Dr. Emily Smith-Greenaway and the association between DTR and emergency department visits in New York City with Dr. Ho. He co-authored a paper led by Dr. Christopher Kyba that highlights the benefit of multiple-angle observations for studying artificial light at night [hyperlink to https://doi.org/10.1029/2021JD036382].

Collaborating with HKU Light Pollution Research [hyperlink to https://nightsky.physics.hku.hk] with Dr. Jason C.S. Pun, he gave several oral presentations on light pollution research at Artificial Light at Night (ALAN) 2021, Light Pollution: Theory, Modeling and Measurements (LPTMM) 2022, ALAN 2023 and the International Geoscience and Remote Sensing Symposium (IGARSS) 2023. He gave an oral presentation about using Zhuhai-1 hyperspectral data to estimate PM2.5 and PM10 concentrations in IGARSS 2022 and two oral presentations in IGARSS 2023 at Pasadena about using multi-source data and citizen science to capture the impacts of Earth Hour 2021 in Hong Kong and nighttime remote sensing data to identify light pollution. He was a session chair in IGARSS 2023.

His research supports the United Nations Sustainable Development Goals (03 Good Health & Well-Being; 13 Climate Action; 15 Life on Land).

R.A., Light Pollution, The University of Hong Kong
B.S., Geographic Information Science, Sun Yat-Sen University, Guangzhou, China

Liu, Shengjie, An-Min Wu, and Hung Chak Ho (2023). Spatial variability of diurnal temperature range and its associations with local climate zone, neighborhood environment and mortality in Los Angeles. Urban Climate 49, 101526.

Kyba, Christopher CM, Martin Aubé, Salvador Bará, Andrea Bertolo, Constantinos A Bouroussis, Stefano Cavazzani, Brian R Espey, Fabio Falchi, Geza Gyuk, Andreas Jechow, Miroslav Kocifaj, Zoltán Kolláth, Héctor Lamphar, Noam Levin, Shengjie Liu, Steven D Miller, Sergio Ortolani, Chun Shing Jason Pun, Salvador José Ribas, Thomas Ruhtz, Alejandro Sánchez de Miguel, Matthias Schneider, Ranjay Man Shrestha, Alexandre Simoneau, Chu Wing So, Tobias Storch, Kai Pong Tong, Diane Turnshek, Ken Walczak, Jun Wang, Zhuosen Wang, and Jianglong Zhang (2022). Multiple Angle Observations Would Benefit Visible Band Remote Sensing Using Night Lights. Journal of Geophysical Research: Atmospheres 127(12), e2021JD036382.

Liu, Shengjie, Zhize Zhou, Huaxiang Ding, Yuanjun Zhong, and Qian Shi (2021). Crop Mapping Using Sentinel Full-Year Dual-Polarized SAR Data and a CPU-Optimized Convolutional Neural Network With Two Sampling Strategies. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing 14, 7017-7031.

Liu, Shengjie, Qian Shi, and Liangpei Zhang (2021). Few-shot Hyperspectral Image Classification with Unknown Classes Using Multitask Deep Learning. IEEE Transactions on Geoscience and Remote Sensing 59(6), 5085–5102.

Liu, Shengjie, Haowen Luo, and Qian Shi (2021). Active Ensemble Deep Learning for Polarimetric Synthetic Aperture Radar Image Classification. IEEE Geoscience and Remote Sensing Letters 18(9), 1580–1584.

Liu, Shengjie, and Qian Shi (2020). Local Climate Zone Mapping as Remote Sensing Scene Classification Using Deep Learning: A Case Study of Metropolitan China. ISPRS Journal of Photogrammetry and Remote Sensing 164, 229-242.

Liu, Shengjie, and Qian Shi (2020). Multitask Deep Learning With Spectral Knowledge for Hyperspectral Image Classification. IEEE Geoscience and Remote Sensing Letters 17(12), 2110-2114.

Liu, Shengjie, Zhixin Qi, Xia Li, and Anthony Gar-On Yeh (2019). Integration of Convolutional Neural Networks and Object Based Post-Classification Refinement for Land Use and Land Cover Mapping with Optical and SAR Data. Remote Sensing 11(6), 690.

Liu, Shengjie, Chu Wing So, Hung Chak Ho, Qian Shi, and Chun Shing Jason Pun (2023). Using high-resolution nighttime remote sensing data to identify light sources in Hong Kong. IGARSS 2023, in press

Liu, Shengjie, Chu Wing So, Xiang Feng Foo, and Chun Shing Jason Pun (2023). Using multisource data to capture the impacts of Earth Hour 2021: A case study of Hong Kong. IGARSS 2023, in press

Liu, Shengjie, and Qian Shi (2022). Estimating PM2.5 and PM10 on Zhuhai-1 hyperspectral imagery. IGARSS 2022, 5933-5936.

Liu, Shengjie, Chu Wing So, and Chun Shing Jason Pun (2021). Analyzing long-term artificial light at night using VIIRS monthly product with land use data: Preliminary result of Hong Kong. IGARSS 2021, 6821-6824.

Liu, Shengjie, and Qian Shi (2021). Multi-label local climate zone mapping as scene classification using very high resolution imagery: Preliminary result of Hong Kong. IGARSS 2021, 6809-6812.

Liu, Shengjie, Haowen Luo, Ying Tu, Zhi He, and Jun Li (2018). Wide contextual residual network with active learning for remote sensing image classification. IGARSS 2018, 7145-7148.

Liu, Shengjie, Chu Wing So, Hung Chak Ho, Qian Shi, and Jason Chun Shing Pun (2023). High inequality of artificial light due to commercial and sports lighting in Hong Kong. ALAN 2023, Calgary, Alberta, Canada, August 2023.

Liu, Shengjie, Chu Wing So, and Chun Shing Jason Pun (2022). Analyzing the sources and variations of night lights between 2012 and 2019 in Hong Kong from VIIRS monthly products. LPTMM 2022, Santiago de Compostela, Galicia, Spain, June 2022.

Liu, Shengjie, Chu Wing So, and Chun Shing Jason Pun (2021). The relationship between night sky brightness and remote sensing data: Preliminary result from Luojia-1 and the International Space Station. ALAN 2021, Lleida, Catalonia, Spain, June 2021.

Session Chair, Hyperspectral Imaging Classification, IGARSS 2023

Member, IEEE Geoscience and Remote Sensing Society (GRSS)

Member, IEEE GRSS Image Analysis and Data Fusion (IADF) Technical Committee

Reviewer for IEEE Geoscience and Remote Sensing Letters (GRSL), IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing (JSTARS), IEEE Transactions on Geoscience and Remote Sensing (TGRS), Knowledge-Based Systems, Pattern Recognition Letters, Remote Sensing Letters, Scientific Reports, Urban Climate