Introspective classification for pedestrian detection

Calum G. Blair, John Thompson, Neil M. Robertson

Research output: Chapter in Book/Report/Conference proceedingConference contribution

3 Citations (Scopus)

Abstract

State-of-the-art pedestrian detectors are capable of finding humans in images with reasonable accuracy. However, accurate object detectors such as Integral Channel Features (ICF) do not provide good reliability; they are unable to identify detections which they are less confident (or more uncertain) about. We apply existing methods for generating probabilistic measures from classifier scores (such as Piatt exponential scaling and Isotonic Regression) and compare these to Gaussian Process classifiers (GPCs), which can provide more informative predictive variance. GPCs are less accurate than ICF classifiers, but GPCs and Adaboost with Piatt scaling both provide improved reliability over existing methods.

Original languageEnglish
Title of host publication2014 Sensor Signal Processing for Defence
PublisherIEEE
Number of pages5
ISBN (Print)9781479952946
DOIs
Publication statusPublished - 31 Oct 2014
Event4th Sensor Signal Processing for Defence 2014 - Edinburgh, Edinburgh, United Kingdom
Duration: 8 Sep 20149 Sep 2014

Conference

Conference4th Sensor Signal Processing for Defence 2014
Abbreviated titleSSPD 2014
CountryUnited Kingdom
CityEdinburgh
Period8/09/149/09/14

ASJC Scopus subject areas

  • Signal Processing
  • Computer Science Applications
  • Electrical and Electronic Engineering

Fingerprint Dive into the research topics of 'Introspective classification for pedestrian detection'. Together they form a unique fingerprint.

  • Cite this

    Blair, C. G., Thompson, J., & Robertson, N. M. (2014). Introspective classification for pedestrian detection. In 2014 Sensor Signal Processing for Defence [6943310] IEEE. https://doi.org/10.1109/SSPD.2014.6943310