Transport Object Detection in Street View Imagery Using Decomposed Convolutional Neural Networks

Cloud Bai, Changjing Shang, Ying Li, Liang Shen, Xianwen Zeng, Qiang Shen

Research output: Contribution to conferencePaperpeer-review

47 Downloads (Pure)

Abstract

Deep learning has achieved great successes in performing many visual recognition tasks including object detection. Nevertheless, existing deep networks are computationally expensive and memory intensive, hindering their deployment in resource-constrained environments, such as mobile or embedded devices that are widely used by city travellers. Recently, estimating city-level travel patterns using street imagery has shown to be a potentially valid way according to a case study with Google Street View (GSV), addressing a critical challenge in transport object detection. This paper presents a compressed deep network using tensor decomposition to detect transport objects in GSV images, which is sustainable and eco-friendly. In particular, a new dataset named Transport Mode Share-Tokyo (TMS-Tokyo) is created to serve the public for transport object detection. This is based on the selection and filtering of 32,555 acquired images that involve 50,827 visible transport objects (including cars, pedestrians, buses, trucks, motors, vans, cyclists and parked bicycles) from the GSV imagery of Tokyo. Then a compressed convolutional neural network (termed SVDet) is proposed for street view object detection via tensor train decomposition on a given baseline detector. Experimental results conducted on the TMS-Tokyo dataset demonstrate that SVDet can achieve promising performance in comparison with conventional deep detection networks.
Original languageEnglish
Number of pages12
Publication statusAccepted/In press - 2022
EventUKCI 2022 - Frederick Mappin Building, University of Sheffield, Sheffield, United Kingdom of Great Britain and Northern Ireland
Duration: 07 Sept 202209 Sept 2022

Conference

ConferenceUKCI 2022
Country/TerritoryUnited Kingdom of Great Britain and Northern Ireland
CitySheffield
Period07 Sept 202209 Sept 2022

Keywords

  • Convolutional Neural Networks,
  • Street-view Object Detection,
  • Tensor Train Decomposition.

Fingerprint

Dive into the research topics of 'Transport Object Detection in Street View Imagery Using Decomposed Convolutional Neural Networks'. Together they form a unique fingerprint.
  • Aberdoc and President Scholarship

    Abu-Bakare, Amal (Recipient), 21 Apr 2016

    Prize: Fellowship awarded competitively

Cite this