Exploring Object-Centric and Scene-Centric CNN Features and Their Complementarity for Human Rights Violations Recognition in Images
Identifying potential abuses of human rights through imagery is a novel and challenging task in the field of computer vision, which will enable to expose human rights violations over large-scale data that may otherwise be impossible. While standard databases for object and scene categorization conta...
Main Authors: | , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
IEEE
2019-01-01
|
Series: | IEEE Access |
Subjects: | |
Online Access: | https://ieeexplore.ieee.org/document/8606079/ |
id |
doaj-e53c996a7457476caabe31776c3733fa |
---|---|
record_format |
Article |
spelling |
doaj-e53c996a7457476caabe31776c3733fa2021-03-29T22:46:33ZengIEEEIEEE Access2169-35362019-01-017100451005610.1109/ACCESS.2019.28917458606079Exploring Object-Centric and Scene-Centric CNN Features and Their Complementarity for Human Rights Violations Recognition in ImagesGrigorios Kalliatakis0https://orcid.org/0000-0002-2194-7709Shoaib Ehsan1https://orcid.org/0000-0001-9631-1898Ales Leonardis2Maria Fasli3Klaus D. McDonald-Maier4School of Computer Science and Electronic Engineering, University of Essex, Colchester, U.K.School of Computer Science and Electronic Engineering, University of Essex, Colchester, U.K.School of Computer Science, University of Birmingham, Birmingham, U.K.School of Computer Science and Electronic Engineering, University of Essex, Colchester, U.K.School of Computer Science and Electronic Engineering, University of Essex, Colchester, U.K.Identifying potential abuses of human rights through imagery is a novel and challenging task in the field of computer vision, which will enable to expose human rights violations over large-scale data that may otherwise be impossible. While standard databases for object and scene categorization contain hundreds of different classes, the largest available dataset of human rights violations contains only four classes. Here, we introduce the human rights archive (HRA) database, a verified-by-experts repository of 3050 human rights violations photographs, labeled with human rights semantic categories, comprising a list of the types of human rights abuses encountered at present. With the HRA dataset and a two-phase transfer learning scheme, we fine-tuned the state-of-the-art deep convolutional neural networks (CNNs) to provide human rights violations classification CNNs. We also present extensive experiments refined to evaluate how well object-centric and scene-centric CNN features can be combined for the task of recognizing human rights abuses. With this, we show that the HRA database poses a challenge at a higher level for the well-studied representation learning methods and provide a benchmark in the task of human rights violations recognition in visual context. We expect that this dataset can help to open up new horizons on creating systems that are able to recognize rich information about human rights violations.https://ieeexplore.ieee.org/document/8606079/Computer visionimage interpretationvisual recognitionconvolutional neural networkshuman rights abuses recognition |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Grigorios Kalliatakis Shoaib Ehsan Ales Leonardis Maria Fasli Klaus D. McDonald-Maier |
spellingShingle |
Grigorios Kalliatakis Shoaib Ehsan Ales Leonardis Maria Fasli Klaus D. McDonald-Maier Exploring Object-Centric and Scene-Centric CNN Features and Their Complementarity for Human Rights Violations Recognition in Images IEEE Access Computer vision image interpretation visual recognition convolutional neural networks human rights abuses recognition |
author_facet |
Grigorios Kalliatakis Shoaib Ehsan Ales Leonardis Maria Fasli Klaus D. McDonald-Maier |
author_sort |
Grigorios Kalliatakis |
title |
Exploring Object-Centric and Scene-Centric CNN Features and Their Complementarity for Human Rights Violations Recognition in Images |
title_short |
Exploring Object-Centric and Scene-Centric CNN Features and Their Complementarity for Human Rights Violations Recognition in Images |
title_full |
Exploring Object-Centric and Scene-Centric CNN Features and Their Complementarity for Human Rights Violations Recognition in Images |
title_fullStr |
Exploring Object-Centric and Scene-Centric CNN Features and Their Complementarity for Human Rights Violations Recognition in Images |
title_full_unstemmed |
Exploring Object-Centric and Scene-Centric CNN Features and Their Complementarity for Human Rights Violations Recognition in Images |
title_sort |
exploring object-centric and scene-centric cnn features and their complementarity for human rights violations recognition in images |
publisher |
IEEE |
series |
IEEE Access |
issn |
2169-3536 |
publishDate |
2019-01-01 |
description |
Identifying potential abuses of human rights through imagery is a novel and challenging task in the field of computer vision, which will enable to expose human rights violations over large-scale data that may otherwise be impossible. While standard databases for object and scene categorization contain hundreds of different classes, the largest available dataset of human rights violations contains only four classes. Here, we introduce the human rights archive (HRA) database, a verified-by-experts repository of 3050 human rights violations photographs, labeled with human rights semantic categories, comprising a list of the types of human rights abuses encountered at present. With the HRA dataset and a two-phase transfer learning scheme, we fine-tuned the state-of-the-art deep convolutional neural networks (CNNs) to provide human rights violations classification CNNs. We also present extensive experiments refined to evaluate how well object-centric and scene-centric CNN features can be combined for the task of recognizing human rights abuses. With this, we show that the HRA database poses a challenge at a higher level for the well-studied representation learning methods and provide a benchmark in the task of human rights violations recognition in visual context. We expect that this dataset can help to open up new horizons on creating systems that are able to recognize rich information about human rights violations. |
topic |
Computer vision image interpretation visual recognition convolutional neural networks human rights abuses recognition |
url |
https://ieeexplore.ieee.org/document/8606079/ |
work_keys_str_mv |
AT grigorioskalliatakis exploringobjectcentricandscenecentriccnnfeaturesandtheircomplementarityforhumanrightsviolationsrecognitioninimages AT shoaibehsan exploringobjectcentricandscenecentriccnnfeaturesandtheircomplementarityforhumanrightsviolationsrecognitioninimages AT alesleonardis exploringobjectcentricandscenecentriccnnfeaturesandtheircomplementarityforhumanrightsviolationsrecognitioninimages AT mariafasli exploringobjectcentricandscenecentriccnnfeaturesandtheircomplementarityforhumanrightsviolationsrecognitioninimages AT klausdmcdonaldmaier exploringobjectcentricandscenecentriccnnfeaturesandtheircomplementarityforhumanrightsviolationsrecognitioninimages |
_version_ |
1724190970119979008 |