Reinforcement-based Display-size Selection for Frugal Satellite Image Change Detection
Not on the map. Why?
Abstract (arXiv)
We introduce a novel interactive satellite image change detection algorithm based on active learning. The proposed method is iterative and consists in frugally probing the user (oracle) about the labels of the most critical images, and according to the oracle's annotations, it updates change detection results. First, we consider a probabilistic framework which assigns to each unlabeled sample a relevance measure modeling how critical is that sample when training change detection functions. We obtain these relevance measures by minimizing an objective function mixing diversity, representativity and uncertainty. These criteria when combined allow exploring different data modes and also refining change detections. Then, we further explore the potential of this objective function, by considering a reinforcement learning approach that finds the best combination of diversity, representativity and uncertainty as well as display-sizes through active learning iterations, leading to better generalization as shown through experiments in interactive satellite image change detection.
Key facts. Cited by 1 paper in the map (data through 7 Oct 2026). Builds on 44 of its 138 references that are in the map. First posted on arXiv 28 Dec 2023.
Builds on
44 of 138 references.
References · 44
- 2017Attention Is All You Needcited by 59,932
- 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 29,915
- 2016Semi-Supervised Classification with Graph Convolutional Networkscited by 11,282
- 2017Inductive Representation Learning on Large Graphscited by 6,284
- 2015Rethinking the Inception Architecture for Computer Visioncited by 5,451
- 2015Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learningcited by 4,416
- 2017Prototypical Networks for Few-shot Learningcited by 3,001
- 2019A Comprehensive Survey on Graph Neural Networkscited by 2,906
- 2016Matching Networks for One Shot Learningcited by 2,246
- 2016Convolutional Neural Networks on Graphs with Fast Localized Spectral Filteringcited by 2,085
- 2016Inception-v4, Inception-ResNet and the Impact of Residual Connections on Learningcited by 1,509
- 2013Spectral Networks and Locally Connected Networks on Graphscited by 1,097
32 more
- 2015Unsupervised Visual Representation Learning by Context Predictioncited by 1,062
- 2017Deep Bayesian Active Learning with Image Datacited by 812
- 2015Gated Graph Sequence Neural Networkscited by 792
- 2019Self-supervised Visual Feature Learning with Deep Neural Networks: A Surveycited by 656
- 2018Deep Visual Domain Adaptation: A Surveycited by 634
- 2019Deep Batch Active Learning by Diverse, Uncertain Gradient Lower Boundscited by 602
- 2017Generative Adversarial Networks: An Overviewcited by 588
- 2018FastGCN: Fast Learning with Graph Convolutional Networks via Importance Samplingcited by 583
- 2018Is Q-learning Provably Efficient?cited by 485
- 2019BatchBALD: Efficient and Diverse Batch Acquisition for Deep Bayesian Active Learningcited by 428
- 2019Learning Loss for Active Learningcited by 335
- 2015Diffusion-Convolutional Neural Networkscited by 268
- 2019Revisiting Self-Supervised Visual Representation Learningcited by 250
- 2015Deep Convolutional Networks on Graph-Structured Datacited by 246
- 2017CayleyNets: Graph Convolutional Neural Networks with Complex Rational Spectral Filterscited by 203
- 2018Adaptive Sampling Towards Fast Graph Representation Learningcited by 180
- 2018GaAN: Gated Attention Networks for Learning on Large and Spatiotemporal Graphscited by 164
- 2018Adaptive Graph Convolutional Neural Networkscited by 155
- 2020Contextual Diversity for Active Learningcited by 139
- 2018Large-Scale Learnable Graph Convolutional Networkscited by 113
- 2017Learning Active Learning from Datacited by 100
- 2018Efficient Active Learning for Image Classification and Segmentation using a Sample Selection and Conditional Generative Adversarial Networkcited by 74
- 2018Constrained Generation of Semantically Valid Graphs via Regularizing Variational Autoencoderscited by 74
- 2017Generative Adversarial Active Learningcited by 69
- 2019Bayesian Batch Active Learning as Sparse Subset Approximationcited by 68
- 2017Automated Scalable Bayesian Inference via Hilbert Coresetscited by 55
- 2018Adversarial Sampling for Active Learningcited by 51
- 2018Active Learning for Deep Object Detectioncited by 49
- 2018Meta-Learning Transferable Active Learning Policies by Deep Reinforcement Learningcited by 26
- 2020DEAL: Deep Evidential Active Learning for Image Classificationcited by 22
- 2021Learning Connectivity with Graph Convolutional Networks for Skeleton-based Action Recognitioncited by 20
- 2018Cost-effective Object Detection: Active Sample Mining with Switchable Selection Criteriacited by 12
Built on by
Citing papers per year: 2026: 1.
1 paper.