Are there any ‘object detectors’ in the hidden layers of CNNs trained to identify objects or scenes?

Ella M Gale, Nicholas Martin, Ryan Blything, Anh Nguyen, Jeffrey S Bowers

Research output: Contribution to journalArticle (Academic Journal)

Abstract

Various methods of measuring unit selectivity have been developed with the aim of better understanding how neural networks work. But the different measures provide divergent estimates of selectivity, and this has led to different conclusions regarding the conditions in which selective object representations are learned and the functional relevance of these representations. In an attempt to better characterize object selectivity, we undertake a comparison of various selectivity measures on a large set of units in AlexNet, including localist selectivity [7], precision [42], class-conditional mean activity selectivity (CCMAS) [22], the human interpretation of activation maximization (AM) images, and standard signal-detection measures. We find that the different measures provide different estimates of object selectivity, with precision and CCMAS measures providing misleadingly high estimates. Indeed, the most selective units had a poor hit-rate or a high false-alarm rate (or both) in object classification, making them poor object detectors. We fail to find any units that are even remotely as selective as the ‘grandmother cell’ units reported in recurrent neural networks. In order to generalize these results, we compared selectivity measures on units in VGG-16 and GoogLeNet trained on the ImageNet or Places-365 datasets that have been described as ‘object detectors’ according to network dissection [43]. Again, we find poor hit-rates and high false-alarm rates for object classification. We conclude that signal-detection measures provide a better assessment of single-unit selectivity compared to common alternative approaches, and that deep convolutional networks of image classification do not learn object detectors in their hidden layers.
Original languageEnglish
Pages (from-to)60-71
Number of pages12
JournalVision Research
Volume176
Early online date8 Aug 2020
DOIs
Publication statusE-pub ahead of print - 8 Aug 2020

Structured keywords

  • Cognitive Science
  • Visual Perception

Keywords

  • grandmother cell
  • object detectors
  • distributed representations

Fingerprint Dive into the research topics of 'Are there any ‘object detectors’ in the hidden layers of CNNs trained to identify objects or scenes?'. Together they form a unique fingerprint.

  • Projects

    M and M

    Bowers, J. S.

    1/09/1731/08/22

    Project: Research, Parent

    Cite this