Multimodal Damage Identification for Humanitarian Computing Data Set
Download: Data Folder, Data Set Description
Abstract: 5879 captioned images (image and text) from social media related to damage during natural disasters/wars, and belong to 6 classes: Fires, Floods, Natural landscape, Infrastructural, Human, Non-damage.
|
|
Data Set Characteristics: |
Multivariate, Text |
Number of Instances: |
5879 |
Area: |
Social |
Attribute Characteristics: |
Integer |
Number of Attributes: |
N/A |
Date Donated |
2018-06-01 |
Associated Tasks: |
Classification |
Missing Values? |
N/A |
Number of Web Hits: |
17675 |
Source:
Hussein Mouzannar, American University of Beirut (hmm46 '@' aub.edu.lb)
Yara Rizk, American University of Beirut (yar01 '@' aub.edu.lb)
Mariette Awad, American University of Beirut (mariette.awad '@' aub.edu.lb)
Data Set Information:
Samples were retrieved from social media posts including Instagram and Twitter.
Attribute Information:
640x640 RGB images and raw text
Relevant Papers:
Hussein Mouzannar, Yara Rizk, and Mariette Awad, 'Damage Identification in
Social Media Posts using Multimodal Deep Learning,' The 15th International
Conference on Information Systems for Crisis Response and Management
(ISCRAM), Rochester, USA, May 20-23, 2018, pp. 529-543.
Hadi S. Jomaa, Yara Rizk, and Mariette Awad, 'Semantic and Visual Cues
for Humanitarian Computing of Natural Disaster Damage Images,' The 12th
International Conference on Signal Image Technology & Internet Systems (SITIS),
Naples, Italy, Nov. 28-Dec. 1 2016.
Citation Request:
Hussein Mouzannar, Yara Rizk, and Mariette Awad, 'Damage Identification in
Social Media Posts using Multimodal Deep Learning,' The 15th International
Conference on Information Systems for Crisis Response and Management
(ISCRAM), Rochester, USA, May 20-23, 2018, pp. 529-543.
|