• Corpus ID: 41845203

Detection of Flooding Events in Social Multimedia and Satellite Imagery using Deep Neural Networks

@inproceedings{Bischke2017DetectionOF,
  title={Detection of Flooding Events in Social Multimedia and Satellite Imagery using Deep Neural Networks},
  author={Benjamin Bischke and Prakriti Bhardwaj and Aman Gautam and Patrick Helber and Damian Borth and Andreas R. Dengel},
  booktitle={MediaEval},
  year={2017}
}
This paper presents the solution of the DFKI-team for the Multimedia Satellite Task at MediaEval 2017. [] Key Method Additionally, we extend existing network architectures for semantic segmentation to incorporate RGB and Infrared (IR) channels into the model. Our results show that IR information is of vital importance for the detection of flooded areas in satellite imagery.
View Paper
50 Citations
Highly Influential Citations
5
Background Citations
26
Methods Citations
21
Results Citations
4

Tables from this paper

50 Citations

Detecting Flooding in Social Media Imagery Using Multimodal Deep Learning
TLDR
This work explores multimodal deep learning approaches to flood identification in social media images by means of image-metadata fusion and finds that CNNor LSTM-based metadata models do not rival the well-performing baseline image model given the modest size of the available metadata-set.
Fusing Visual Features and Metadata to Detect Flooding in Flickr Images
TLDR
A novel backpropagation technique, Direct Back Propagation (DBP) is used to train a neural network for the classification of flood images in the MediaEval 2017 dataset, showing that the fusion methods improve the classification accuracy compared to their individual counterparts.
Ensembled Convolutional Neural Network Models for Retrieving Flood Relevant Tweets
TLDR
A supervised learning method with an ensemble of 10 Convolutional Neural Networks was applied to classify the tweets in the benchmark to present a method proposed by team EVUSikg for the MediaEval 2018 challenge on Multimedia Satellite Task.
AI-based flood event understanding and quantification using online media and satellite data
TLDR
This work aims to detect floods and also give relevant information about the flood situation such as the water level and the extension of the flooded regions, as specified in the three subtasks, for which of them a specific solution is proposed.
Deep Learning Approaches for Flood Classification and Flood Aftermath Detection
This paper presents the method proposed by team UTAOS for MediaEval 2018 Multimedia Satellite Task: Emergency Response for Flooding Events. In the first challenge, we mainly rely on object and scene…
VRBagged-Net: Ensemble Based Deep Learning Model for Disaster Event Classification
TLDR
The VRBagged-Net with a slight modification also ranked first in the flood-related Multimedia Task at the MediaEval Workshop 2020 and exceeds the previous best result by producing an F1 evaluation of 93.76.
Assessing flood severity from georeferenced photos
TLDR
This paper evaluates the use of deep convolutional neural networks for discriminating images showing direct evidence of a flood, and estimating the severity of the flooding event, concluding that these models for image classification can achieve a very high accuracy on both tasks.
Flood Detection in Social Media Images using Visual Features and Metadata
TLDR
The results demonstrate that the combination of two types of visual features have a positive impact on distinguishing natural disaster images and demonstrate that although the visual feature approach outperforms metadata approach, metadata have certain capability to classify these images.
A Context-Aware Late-Fusion Approach for Disaster Image Retrieval from Social Media
TLDR
This research work introduces a context-aware late-fusion approach for disaster image retrieval from social media, developed a method for image-text content synchronization and spatial-temporal-context event confirmation, and evaluated the role of using different types of features extracted from internal and external data sources.
People and Vehicles in Danger - A Fire and Flood Detection System in Social Media
TLDR
The proposed framework recruits State-of-the-Art deep learning technologies so as to solve a series of image processing and machine learning challenges and provides a near real-time localization solution for detecting and scoring severity safety levels of people and vehicles in flood and fire images.
...
...

References

SHOWING 1-10 OF 13 REFERENCES
The Multimedia Satellite Task at MediaEval 2018
TLDR
The MediaEval 2018 Multimedia Satellite Task focuses on flooding events and consists of two subtasks that seek to go beyond state-of-the-art flooding map generation by focusing on information about road passability and the accessibility of urban infrastructure.
Contextual Enrichment of Remote-Sensed Events with Social Media Streams
TLDR
This work presents a scalable system for the contextual enrichment of satellite images by crawling and analyzing multimedia content from social media and uses Twitter as main data source and analyzes its textual, visual, temporal, geographical and social dimensions.
Overview of ImageCLEF 2017 Population Estimation (Remote) task
TLDR
This paper presents the results that were obtained by the participants of the Population Estimation (or Remote) task as well as a brief summary of some of the approaches that were used.
EuroSAT: A Novel Dataset and Deep Learning Benchmark for Land Use and Land Cover Classification
TLDR
A patch-based land use and land cover classification approach using Sentinel-2 satellite images that covers 13 spectral bands and is comprised of ten classes with a total of 27 000 labeled and geo-referenced images is presented.
Introducing the Sky and the Social Eye
TLDR
The Sky and the Social Eye task was run for the first time as a Grand Challenge at the 2016 ACM Multimedia Conference and as a Task Force at the2016 MediaEval Workshop to create a richer user experience.
Very Deep Convolutional Networks for Large-Scale Image Recognition
TLDR
This work investigates the effect of the convolutional network depth on its accuracy in the large-scale image recognition setting using an architecture with very small convolution filters, which shows that a significant improvement on the prior-art configurations can be achieved by pushing the depth to 16-19 weight layers.
Deep Residual Learning for Image Recognition
TLDR
This work presents a residual learning framework to ease the training of networks that are substantially deeper than those used previously, and provides comprehensive empirical evidence showing that these residual networks are easier to optimize, and can gain accuracy from considerably increased depth.
Deep Cross Residual Learning for Multitask Visual Recognition
TLDR
It is shown how cross-residual learning (CRL) can be integrated in multitask networks to jointly train and detect visual concepts across several tasks to achieve competitive, or even better, detection performance on a visual sentiment concept detection problem normally requiring multiple specialized single-task networks.
Large-scale visual sentiment ontology and detectors using adjective noun pairs
TLDR
This work presents a method built upon psychological theories and web mining to automatically construct a large-scale Visual Sentiment Ontology (VSO) consisting of more than 3,000 Adjective Noun Pairs (ANP) and proposes SentiBank, a novel visual concept detector library that can be used to detect the presence of 1,200 ANPs in an image.
DeepSentiBank: Visual Sentiment Concept Classification with Deep Convolutional Neural Networks
TLDR
Performance evaluation shows the newly trained deep CNNs model SentiBank 2.0 (or called DeepSentiBank) is significantly improved in both annotation accuracy and retrieval performance, compared to its predecessors which mainly use binary SVM classification models.
...
...