When Naive Bayes Nearest Neighbors Meet Convolutional Neural Networks

Ilja Kuzborskij, Fabio Maria Carlucci, Barbara Caputo; Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2016, pp. 2100-2109

Abstract


Since Convolutional Neural Networks (CNNs) have become the leading learning paradigm in visual recognition, Naive Bayes Nearest Neighbor (NBNN)-based classifiers have lost momentum in the community. This is because (1) such algorithms cannot use CNN activations as input features; (2) they cannot be used as final layer of CNN architectures for end-to-end training , and (3) they are generally not scalable and hence cannot handle big data. This paper proposes a framework that addresses all these issues, thus bringing back NBNNs on the map. We solve the first by extracting CNN activations from local patches at multiple scale levels, similarly to [13]. We address simultaneously the second and third by proposing a scalable version of Naive Bayes Non-linear Learning (NBNL, [7]). Results obtained using pre-trained CNNs on standard scene and domain adaptation databases show the strength of our approach, opening a new season for NBNNs.

Related Material


[pdf] [supp]
[bibtex]
@InProceedings{Kuzborskij_2016_CVPR,
author = {Kuzborskij, Ilja and Carlucci, Fabio Maria and Caputo, Barbara},
title = {When Naive Bayes Nearest Neighbors Meet Convolutional Neural Networks},
booktitle = {Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR)},
month = {June},
year = {2016}
}