When Deep Learning Meets Metric Learning: Remote Sensing Image Scene Classification via Learning Discriminative CNNs

By Gong Cheng, Ceyuan Yang, Xiwen Yao, Lei Guo, Junwei Han. IEEE Transactions on Geoscience and Remote Sensing, 2018

1,041 citations recorded by the source corpus.

Subfields: Atmospheric Science, Computer Vision and Pattern Recognition, Media Technology

View Paper: https://www.semanticscholar.org/paper/0f78ad40306f7e6bb4d609fa9463a085cfba6f02

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Atmospheric Science

1. Zhenwei Shi
2. Liangpei Zhang
3. Marco Körner
4. Marc Rußwurm
5. Gong Cheng
6. Junwei Han
7. Guisong Xia
8. Lichao Mou

In Computer Vision and Pattern Recognition

1. Junwei Han
2. Gong Cheng
3. Peicheng Zhou
4. Guisong Xia
5. Zhenwei Shi
6. Liangpei Zhang
7. Lei Guo
8. Ke Li

In Media Technology

1. Gong Cheng
2. Junwei Han
3. Zhenwei Shi
4. Liangpei Zhang
5. Peicheng Zhou
6. Guisong Xia
7. Marco Körner
8. Lei Guo

Citations in the corpus

• A Deeply Supervised Attention Metric-Based Network and an Open Aerial Image Dataset for Remote Sensing Change Detection (2021, methodology): “Of late, deep metric learning [28] has also made significant progress in the area of CD [29], [30].”
• Self-Supervised Learning in Remote Sensing: A review (2022, background): “NEGATIVE SAMPLING IN REMOTE SENSING While contrastive SSL is relatively new, the wide use of contrastive loss in remote sensing [178]–[183] can date back to [184], where the authors imposed a supervised contrastive regularization term on the CNN features for remote sensing scene classification.”
• Research Progress on Few-Shot Learning for Remote Sensing Image Interpretation (2021, methodology): “Cheng et al. [90] introduced a metric learning regularization term on the CNN features to make the model more discriminative and verified that the model in this method can obtain more information from the training data of smaller size.”
• Remote Sensing Image Scene Classification Meets Deep Learning: Challenges, Methods, Benchmarks, and Opportunities (2020, methodology): “To tackle the challenges, some representative methods [72], [73], [145] have been introduced over the past few years, such as multitask learning (e.”
• Deep Unsupervised Embedding for Remotely Sensed Images Based on Spatially Augmented Momentum Contrast (2020, background): “One common strategy adopted in the literature is based on sampling the negative patches within each minibatch [39].”
• Object Detection in Optical Remote Sensing Images: A Survey and A New Benchmark (2019, background): “…recently, deep learning based algorithms have been dominating the top accuracy benchmarks for various visual recognition tasks (Chen et al., 2018; Cheng et al., 2018a; Clément et al., 2013; Ding et al., 2017; Hinton et al., 2012; Hou et al., 2017; Krizhevsky et al., 2012; Mikolov et al., 2012;…”
• Classification of Remote Sensing Images Using EfficientNet-B3 CNN Model With Attention (2021, background): “A large amount of works that deals with this problem is published in the RS literature [9]–[29].”
• Vision Transformers for Remote Sensing Image Classification (2021, background): “Cheng et al. [34] proposed a metric learning regularization on discriminative CNNs features to optimize a new discriminative objective function to make the model more discriminative.”
• Self-Attention for Raw Optical Satellite Time Series Classification (2019, methodology): “Rotation-invariant feature extraction [51] for scene classification [52] and metric learning [53] has been similarly tested for remote sensing image classification.”
• Learning Rotation-Invariant and Fisher Discriminative Convolutional Neural Networks for Object Detection (2019, positive): “To further demonstrate the effectiveness of our proposed RIFD-CNN method, besides object detection, we extend our method to scene classification using remote sensing images [64], [72], which aims to classify remote sensing images into a set of semantic classes according to the image contents.”
• A Spatial-Temporal Attention-Based Method and a New Dataset for Remote Sensing Image Change Detection (2020, background): “During the last few years, deep metric learning has been applied in many remote sensing applications [27,28,60,61].”
• Deep Learning Meets SAR: Concepts, models, pitfalls, and perspectives (2020, background): “Furthermore, specific metric learning approaches to enhance class separability [94] can be adopted for SAR imagery in order to improve the overall classification accuracy.”