DeepGlobe 2018: A Challenge to Parse the Earth through Satellite Images
arXiv:1805.06561 · doi:10.1109/CVPRW.2018.00031
Abstract
We present the DeepGlobe 2018 Satellite Image Understanding Challenge, which includes three public competitions for segmentation, detection, and classification tasks on satellite images. Similar to other challenges in computer vision domain such as DAVIS and COCO, DeepGlobe proposes three datasets and corresponding evaluation methodologies, coherently bundled in three competitions with a dedicated workshop co-located with CVPR 2018. We observed that satellite imagery is a rich and structured source of information, yet it is less investigated than everyday images by computer vision researchers. However, bridging modern computer vision with remote sensing data analysis could have critical impact to the way we understand our environment and lead to major breakthroughs in global urban planning or climate change research. Keeping such bridging objective in mind, DeepGlobe aims to bring together researchers from different domains to raise awareness of remote sensing in the computer vision community and vice-versa. We aim to improve and evaluate state-of-the-art satellite image understanding approaches, which can hopefully serve as reference benchmarks for future research in the same topic. In this paper, we analyze characteristics of each dataset, define the evaluation criteria of the competitions, and provide baselines for each task.
Dataset description for DeepGlobe 2018 Challenge at CVPR 2018
References in corpus (2)
Cited by in corpus (100)
- A Survey on Scalable LoRaWAN for Massive IoT: Recent Advances, Potentials, and Challenges
- RSVQA: Visual Question Answering for Remote Sensing Data
- MAP-Net: Multi Attending Path Neural Network for Building Footprint Extraction from Remote Sensed Imagery
- X-ModalNet: A Semi-Supervised Deep Cross-Modal Network for Classification of Remote Sensing Data
- S2Looking: A Satellite Side-Looking Dataset for Building Change Detection
- Global and Local Contrastive Self-Supervised Learning for Semantic Segmentation of HR Remote Sensing Images
- Stagewise Unsupervised Domain Adaptation with Adversarial Self-Training for Road Segmentation of Remote Sensing Images
- Understanding urban landuse from the above and ground perspectives: a deep learning, multimodal solution
- LoveDA: A Remote Sensing Land-Cover Dataset for Domain Adaptive Semantic Segmentation
- Dense Dilated Convolutions Merging Network for Land Cover Classification
- Aerial Imagery for Roof Segmentation: A Large-Scale Dataset towards Automatic Mapping of Buildings
- RescueNet: A High Resolution UAV Semantic Segmentation Benchmark Dataset for Natural Disaster Damage Assessment
- DDU-Net: Dual-Decoder-U-Net for Road Extraction Using High-Resolution Remote Sensing Images
- Enabling Country-Scale Land Cover Mapping with Meter-Resolution Satellite Imagery
- Semantics for Robotic Mapping, Perception and Interaction: A Survey
- Self-supervised remote sensing feature learning: Learning Paradigms, Challenges, and Future Works
- A Comprehensive Analysis of Weakly-Supervised Semantic Segmentation in Different Image Domains
- SpaceNet MVOI: a Multi-View Overhead Imagery Dataset
- DiResNet: Direction-aware Residual Network for Road Extraction in VHR Remote Sensing Images
- Semi-Supervised Semantic Segmentation in Earth Observation: The MiniFrance Suite, Dataset Analysis and Multi-task Network Study
- RSAM-Seg: A SAM-based Approach with Prior Knowledge Integration for Remote Sensing Image Semantic Segmentation
- xBD: A Dataset for Assessing Building Damage from Satellite Imagery
- Correcting rural building annotations in OpenStreetMap using convolutional neural networks
- City-Scale Road Extraction from Satellite Imagery v2: Road Speeds and Travel Times
- Cross-dimensional transfer learning in medical image segmentation with deep learning
- From Satellite Imagery to Disaster Insights
- Self-supervised Audiovisual Representation Learning for Remote Sensing Data
- MKANet: A Lightweight Network with Sobel Boundary Loss for Efficient Land-cover Classification of Satellite Remote Sensing Imagery
- Road Extraction with Satellite Images and Partial Road Maps
- URBAN-i: From urban scenes to mapping slums, transport modes, and pedestrians in cities using deep learning and computer vision
- Beyond Geo-localization: Fine-grained Orientation of Street-view Images by Cross-view Matching with Satellite Imagery with Supplementary Materials
- The Synthinel-1 dataset: a collection of high resolution synthetic overhead imagery for building segmentation
- Multi-Modal and Multi-Resolution Data Fusion for High-Resolution Cloud Removal: A Novel Baseline and Benchmark
- False: False Negative Samples Aware Contrastive Learning for Semantic Segmentation of High-Resolution Remote Sensing Image
- Hidden Path Selection Network for Semantic Segmentation of Remote Sensing Images
- Vision-Language Modeling Meets Remote Sensing: Models, Datasets and Perspectives
- Adaptive Linear Span Network for Object Skeleton Detection
- Birds of A Feather Flock Together: Category-Divergence Guidance for Domain Adaptive Segmentation
- Agriculture-Vision: A Large Aerial Image Database for Agricultural Pattern Analysis
- GeoGAN: A Conditional GAN with Reconstruction and Style Loss to Generate Standard Layer of Maps from Satellite Images
- Meta-Learning for Few-Shot Land Cover Classification
- Learning to Downsample for Segmentation of Ultra-High Resolution Images
- Deploying machine learning to assist digital humanitarians: making image annotation in OpenStreetMap more efficient
- HazeSpace2M: A Dataset for Haze Aware Single Image Dehazing
- SegBlocks: Block-Based Dynamic Resolution Networks for Real-Time Segmentation
- PP-LinkNet: Improving Semantic Segmentation of High Resolution Satellite Imagery with Multi-stage Training
- Multitask Learning for Large-scale Semantic Change Detection
- FAIR1M: A Benchmark Dataset for Fine-grained Object Recognition in High-Resolution Remote Sensing Imagery
- UAVid: A Semantic Segmentation Dataset for UAV Imagery
- NL-LinkNet: Toward Lighter but More Accurate Road Extraction with Non-Local Operations
- Semantic Labeling of Large-Area Geographic Regions Using Multi-View and Multi-Date Satellite Images and Noisy OSM Training Labels
- Scribble-based Weakly Supervised Deep Learning for Road Surface Extraction from Remote Sensing Images
- SustainBench: Benchmarks for Monitoring the Sustainable Development Goals with Machine Learning
- Population Mapping in Informal Settlements with High-Resolution Satellite Imagery and Equitable Ground-Truth
- On Creating Benchmark Dataset for Aerial Image Interpretation: Reviews, Guidances and Million-AID
- CNN-Based Semantic Change Detection in Satellite Imagery
- Randomized Histogram Matching: A Simple Augmentation for Unsupervised Domain Adaptation in Overhead Imagery
- Neural Turtle Graphics for Modeling City Road Layouts
- Deep Learning-based Aerial Image Segmentation with Open Data for Disaster Impact Assessment
- Revisiting Classical Bagging with Modern Transfer Learning for On-the-fly Disaster Damage Detector
- Magnifier: A Multi-grained Neural Network-based Architecture for Burned Area Delineation
- Model Generalization in Deep Learning Applications for Land Cover Mapping
- Effective Data Fusion with Generalized Vegetation Index: Evidence from Land Cover Segmentation in Agriculture
- Satellite-Net: Automatic Extraction of Land Cover Indicators from Satellite Imagery by Deep Learning
- IRSAMap:Towards Large-Scale, High-Resolution Land Cover Map Vectorization
- Boundary Regularized Building Footprint Extraction From Satellite Images Using Deep Neural Network
- Robust object extraction from remote sensing data
- Discoverability in Satellite Imagery: A Good Sentence is Worth a Thousand Pictures
- Robust Deep Learning with Active Noise Cancellation for Spatial Computing
- SRMF: A Data Augmentation and Multimodal Fusion Approach for Long-Tail UHR Satellite Image Segmentation
- Financial Table Extraction in Image Documents
- Foveation for Segmentation of Ultra-High Resolution Images
- Disaster mapping from satellites: damage detection with crowdsourced point labels
- Toward task-driven satellite image super-resolution
- Textual and Visual Guided Task Adaptation for Source-Free Cross-Domain Few-Shot Segmentation
- TerrAInav Sim: An Open-Source Simulation of UAV Aerial Imaging from Satellite Data
- GridTracer: Automatic Mapping of Power Grids using Deep Learning and Overhead Imagery
- Map Generation from Large Scale Incomplete and Inaccurate Data Labels
- Domain-Adversarial Training of Self-Attention Based Networks for Land Cover Classification using Multi-temporal Sentinel-2 Satellite Imagery
- MultiNet: Segmenting Flooded Buildings via Fusion of Multiresolution, Multisensor, and Multitemporal Satellite Imagery
- Context-aware Padding for Semantic Segmentation
- SkyScapes -- Fine-Grained Semantic Understanding of Aerial Scenes
- Bidirectional Multi-scale Attention Networks for Semantic Segmentation of Oblique UAV Imagery
- Scale Aware Adaptation for Land-Cover Classification in Remote Sensing Imagery
- SegAssess: Panoramic quality mapping for robust and transferable unsupervised segmentation assessment
- Collaborative Global-Local Networks for Memory-Efficient Segmentation of Ultra-High Resolution Images
- AI Powered Road Network Prediction with Multi-Modal Data
- Towards Reliable Evaluation of Road Network Reconstructions
- Quantization in Relative Gradient Angle Domain For Building Polygon Estimation
- DDU-Net: A Domain Decomposition-Based CNN for High-Resolution Image Segmentation on Multiple GPUs
- Beyond Road Extraction: A Dataset for Map Update using Aerial Images
- LUAI Challenge 2021 on Learning to Understand Aerial Images
- Mask Clustering-based Annotation Engine for Large-Scale Submeter Land Cover Mapping
- Deep Built-Structure Counting in Satellite Imagery Using Attention Based Re-Weighting
- Generating a Training Dataset for Land Cover Classification to Advance Global Development
- Learning Geocentric Object Pose in Oblique Monocular Images
- Heterogeneous Grid Convolution for Adaptive, Efficient, and Controllable Computation
- Crowd-Sourced Road Quality Mapping in the Developing World
- Addressing the Invisible: Street Address Generation for Developing Countries with Deep Learning
- Dense Fusion Classmate Network for Land Cover Classification