A5067888699- Human Pose and Action Recognition
- Video Surveillance and Tracking Methods
- Flood Risk Assessment and Management
- Hydrology and Watershed Management Studies
- Remote Sensing in Agriculture
- AI and Multimedia in Education
- Anomaly Detection Techniques and Applications
- Gait Recognition and Analysis
- Groundwater and Watershed Analysis
- Fire effects on ecosystems
- Remote Sensing and Land Use
- Media Studies and Communication
- Land Use and Ecosystem Services
- Water resources management and optimization
- Human Motion and Animation
- Remote-Sensing Image Classification
- Leaf Properties and Growth Measurement
- Plant Water Relations and Carbon Dynamics
- Machine Learning and ELM
- Geochemistry and Geologic Mapping
- Water Resources and Management
- Atmospheric and Environmental Gas Dynamics
- Advanced Technologies in Various Fields
- Hand Gesture Recognition Systems
- Environmental Changes in China
Nanjing Institute of Geography and Limnology
2018-2023
Chinese Academy of Sciences
2018-2023
University of Hong Kong
2020-2023
Hohai University
2020-2023
Ministry of Water Resources of the People's Republic of China
2021-2022
Chinese University of Hong Kong
2020
Nanjing University of Information Science and Technology
2018
Abstract Background ‘Megafire’ is an emerging concept commonly used to describe fires that are extreme in terms of size, behaviour, and/or impacts, but the term’s meaning remains ambiguous. Approach We sought resolve ambiguity surrounding ‘megafire’ by conducting a structured review use and definition term several languages peer‐reviewed scientific literature. collated definitions descriptions megafire identified criteria frequently invoked define megafire. recorded size location megafires...
Abstract Rivers are among the most diverse, dynamic, and productive ecosystems on Earth. River flow regimes constantly changing, but characterizing understanding such changes have been challenging from a long-term global perspective. By analyzing water extent variations observed four-decade Landsat imagery, we here provide attribution of recent in river regime to morphological dynamics (e.g., channel shifting anabranching), expansion induced by new dams, hydrological signals widening...
In order to satisfy the needs of constant economic growth, pressure exploit natural resources has been increasing rapidly in China. Particularly with implementation National Western Development Strategies since 1999, more and mining activities related infrastructure constructions have conducted on Tibetan Plateau (TP). Mining are known substantial impacts plant dynamics hence water energy cycles. Identifying quantifying their effects vegetation cover critical monitoring protection pristine...
Land use and land cover (LULC) is a key variable of the Earth’s system has become an important indicator to evaluate impact human activities on ecosystems. With increasing demand mine resources, widespread opencast mining led significant changes in LULC caused substantial damage environment. An efficient approach detecting at large scales critical importance mitigating their potential impacts downstream settlements assessing characteristics. In this study, we present novel for enabling...
The number of reservoirs is rapidly increasing owing to the growth world’s economy and related energy water needs. Yet, for vast majority around world, their locations information, especially newly dammed reservoirs, are not readily available due financial, political, or legal considerations. This study proposes an automated method identifying from time series MODIS-derived NDWI (normalized difference index) images. Its main idea lies in detection abrupt changes that associated with...
Point cloud sequence-based 3D action recognition achieves impressive performance and efficiency. Conventional approaches for modeling point sequences usually perform cross-frame spatio-temporal local encoding, thus resulting in intensive computation mutual interference between spatial temporal information extracting. In this work, to avoid we propose a strong parallelized sequence network referred as HyperpointNet recognition. is composed of two serial modules, i.e., hyperpoint embedding...
Human action recognition is an active research area in computer vision. Aiming at the lack of spatial muti-scale information for human recognition, we present a novel framework to recognize actions from depth video sequences using multi-scale Laplacian pyramid motion images (LP-DMI). Each frame projected onto three orthogonal Cartesian planes. Under views, generate (DMI) and construct pyramids as structured feature maps which enhances dynamic motions reduces redundant static bodies. We...
Establishing protected areas (PAs) in Amazon forests is crucial for safeguarding tropical forest ecosystem from human land use and mitigating degradation. However, PAs across the basin have increasingly suffered intensified fires. Understanding post-fire recovery trajectories these essential assessing resilience effectiveness of PAs. under natural conditions remain unclear, as settlements often disrupt or influence process, potentially diminishing rates potential. To address this challenge,...
Human action recognition (HAR) as the most representative human-centred computer vision task is critical in human resource management (HRM), especially recruitment, performance appraisal, and employee training. Currently, prevailing approaches to primarily emphasize either temporal or spatial features while overlooking intricate interplay between these two dimensions. This oversight leads less precise robust classification within complex recruitment environments. In this paper, we propose a...
More and more attention has been focused on the human action recognition domain. The existing methods are mostly based single-mode data. However, data lacks adequate information. So, it is necessary to propose multimode In this paper, we extract two kinds of features from depth videos skeleton sequences, named STDMI-HOG STjoint feature respectively. extracted a new map Spatial-Temporal Depth Motion Image by Histogram Oriented Gradient. sequences ST-GCN extractor. Then connected make up...
Abstract. Due to the implementation of national policy, desertification in Ningxia has been gradually reduced, but overall situation is still serious. Rainfall Use Efficiency(RUE) can make some improvement problem that precipitation a great influence on vegetation arid area and fully reflect dynamic characteristics desertification. Using MOD13Q1 data, land use classification map, as well non-remote sensing data such meteorological social statistics paper carries out evaluation status based...
Human action recognition is an active research area in computer vision. Although great process has been made, previous methods mostly recognize actions based on depth data at only one scale, and thus they often neglect multi-scale features that provide additional information practical application scenarios. In this paper, we present a novel framework focusing motion to human from video sequences. We propose feature map called Laplacian pyramid images(LP-DMI). employ images (DMI) as the...