Fighting selection bias in statistical learning: application to visual recognition from biased image databases

Abstract

AbstractIn practice, and especially when training deep neural networks, visual recognition rules are often learned based on various sources of information. On the other hand, the recent deployment of facial recognition systems with uneven performances on different population segments has highlighted the representativeness issues induced by a naive aggregation of the datasets. In this paper, we show how biasing models can remedy these problems. Based on the (approximate) knowledge of the biasing mechanisms at work, our approach consists in reweighting the observations, so as to form a nearly debiased estimator of the target distribution. One key condition is that the supports of the biased distributions must partly overlap, and cover the support of the target distribution. In order to meet this requirement in practice, we propose to use a low dimensional image representation, shared across the image databases. Finally, we provide numerical experiments highlighting the relevance of our approach.Keywords: Sampling biasselection effectvisual recognitionreliable statistical learning Disclosure statementNo potential conflict of interest was reported by the author(s).Additional informationFundingThis work was partially supported by the research chair ‘Good In Tech : Rethinking innovation and technology as drivers of a better world for and by humans’, under the auspices of the ‘Fondation du Risque’ and in partnership with the Institut Mines-Télécom, Sciences Po, Afnor, Ag2r La Mondiale, CGI France, Danone and Sycomore.