IEEE JSTARS 2022

159 papers
The 2022 edition of the IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing featured research focused on advancing deep learning methodologies for complex geospatial data. A significant portion of the work addressed challenges in semantic labeling and segmentation, utilizing techniques such as calibrated focal loss and transformation consistency regularization for semi-supervised learning. Researchers also explored open-world recognition and out-of-distribution detection, aiming to improve the robustness of neural networks when encountering novel categories or semantic shifts in remote sensing imagery. Other prominent themes included spatiotemporal prediction and change detection. Notable studies utilized convolutional gated recurrent units (ConvGRU) for modeling vegetation dynamics and axial cross-attention mechanisms for bibranch fusion in change detection tasks. The year’s contributions also highlighted distributed deep learning frameworks for large-scale object detection and the integration of spatial disruption for feature disentanglement. Prolific contributors included Binyu Chen, Min Xia, and Ming Qian, among others, who focused on enhancing the spatial and semantic accuracy of remote sensing analysis through innovative neural network architectures.

Top Research Directions

Research directions most represented by IEEE JSTARS 2022 papers.

Mitigating Speckle Noise in Coherent Imaging Systems 5
Overcoming High Intra-Class Variance and Low Inter-Class Similarity 2
Eliminating Spectral Redundancy in Hyperspectral Image Processing 2
Sensing Natural Hazards Through Upper Atmospheric Perturbations 1
Latitudinal Variability of the Equatorial Ionospheric Anomaly 1
Semantic Change Detection for Land Cover Transitions 1
Spatial Neighborhood Consensus for Pseudo Label Correction 1
Decomposing Radar Backscatter into Physical Scattering Mechanisms 1
Deep Learning for Synthetic Aperture Radar Target Recognition 1
Adaptive Spatial Partitioning for Visual Computing 1
Density Aware Initialization for Centroid Based Clustering 1
Balancing Closed-Set Accuracy and Open-Set Rejection 1
Isolating Anomalies through Residual Time Series Decomposition 1
Shuffling Image Patches to Eliminate Semantic Bias 1
Theoretical Foundations and Mechanisms of Knowledge Distillation 1

Top Authors

AuthorPapers
Huiyu Zhou6
Danfeng Hong5
Jocelyn Chanussot5
Licheng Jiao4
Pengfeng Xiao3
Quanhua Liu3
Xueliang Zhang3
Yunsong Li3
Zhenwei Shi3
Adil Khan2
Amir Hussain2
Caleb Robinson2
David A. Clausi2
Jiansheng Chen2
Kevin Garrett2
Leyuan Fang2
Liangpei Zhang2
Lianru Gao2
Ming Qian2
Pedram Ghamisi2

Impactful Papers

Papers from IEEE JSTARS 2022 that shaped research clusters.

Influential

Papers

Title
The Outcome of the 2022 Landslide4Sense Competition: Advanced Landslide Detection From Multisource Satellite Imagery
The Spatio-Temporal Reconstruction of Lake Water Levels Using Deep Learning Models: A Case Study on Altai Mountains
Thin Cloud Removal Fusing Full Spectral and Spatial Features for Sentinel-2 Imagery
TransCloudSeg: Ground-Based Cloud Image Segmentation With Transformer
Transferring Transformer-Based Models for Cross-Area Building Extraction From Remote Sensing Images
Unsupervised Balanced Hash Codes Learning With Multichannel Feature Fusion
Vision Transformer With Contrastive Learning for Remote Sensing Image Scene Classification
Which CAM is Better for Extracting Geographic Objects? A Perspective From Principles and Experiments
WH-MAVS: A Novel Dataset and Deep Learning Benchmark for Multiple Land Use and Land Cover Applications

« prev Page 4/4