SLNSW-UTS: A Historical Image Dataset for Image Multi-Labeling and Retrieval

Junjie Zhang, Jian Zhang, Jianfeng Lu, Chunhua Shen, Kate Curr, Robin Phua, Richard Neville, Elise Edmonds

Research output: Chapter in Book or Report/Conference proceedingConference Proceedingpeer-review

2 Citations (Scopus)

Abstract

This paper introduces a dataset of historical images created by the State Library of New South Wales and the University of Technology Sydney (UTS). The dataset has a total of 29713 images with 119 unique labels. Each image contains multiple labels. We use a CNN-based framework to explore the feasibility of our dataset in image multi-labeling and retrieval research, and extract semantic level image features for future research use. The experiment results illustrate that effective deep learning models can be trained on our dataset. We also introduce five applications that can be studied on our historical image dataset.

Original languageEnglish
Title of host publication2016 International Conference on Digital Image Computing
Subtitle of host publicationTechniques and Applications, DICTA 2016
EditorsAlan Wee-Chung Liew, Jun Zhou, Yongsheng Gao, Zhiyong Wang, Clinton Fookes, Brian Lovell, Michael Blumenstein
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9781509028962
DOIs
Publication statusPublished - 22 Dec 2016
Externally publishedYes
Event2016 International Conference on Digital Image Computing: Techniques and Applications, DICTA 2016 - Gold Coast, Australia
Duration: 30 Nov 20162 Dec 2016

Publication series

Name2016 International Conference on Digital Image Computing: Techniques and Applications, DICTA 2016

Conference

Conference2016 International Conference on Digital Image Computing: Techniques and Applications, DICTA 2016
Country/TerritoryAustralia
CityGold Coast
Period30/11/162/12/16

Keywords

  • dataset construction
  • historical image
  • multi-labeling
  • retrieval

Fingerprint

Dive into the research topics of 'SLNSW-UTS: A Historical Image Dataset for Image Multi-Labeling and Retrieval'. Together they form a unique fingerprint.

Cite this