Automatic Data Cleaning System for Large-Scale Location Image Databases Using a Multilevel Extractor and Multiresolution Dissimilarity Calculation

Hsu Yung Cheng, Chih Chang Yu

Research output: Contribution to journalArticlepeer-review

2 Scopus citations

Abstract

In this article, we propose a system for automatically classifying and cleaning location images in large-scale image databases uploaded by arbitrary users. Detecting incorrect scenes uploaded by users and maintaining the correctness of the database through automatic data cleaning are essential because human inspection is not feasible for verifying massive amounts of data. In this study, we compared different feature extractors using deep convolutional neural networks trained using big data. We designed a multilevel extractor to improve feature extraction. Moreover, a detector based on multiresolution dissimilarity calculation was designed to overcome the issue of large intraclass distances and successfully identify incorrect scenes. The proposed system was validated using a highly challenging dataset with 138,000 images collected from Google Places. The experiments show that the multilevel extractor and the detector based on multiresolution dissimilarity calculation can improve the accuracy in identifying incorrect scenes and achieve satisfying data cleaning results.

Original languageEnglish
Pages (from-to)49-56
Number of pages8
JournalIEEE Intelligent Systems
Volume36
Issue number5
DOIs
StatePublished - 2021

Keywords

  • Automatic data cleaning
  • deep learning
  • location image analysis

Fingerprint

Dive into the research topics of 'Automatic Data Cleaning System for Large-Scale Location Image Databases Using a Multilevel Extractor and Multiresolution Dissimilarity Calculation'. Together they form a unique fingerprint.

Cite this