Cross-Dimensional Weighting for Aggregated Deep Convolutional Features

Yannis Kalantidis, Clayton Mellina, Simon Osindero

ECCVW 2016 pp. 685-701

doi:10.1007/978-3-319-46604-0_48 /eccvw/2016/kalantidis2016eccvw-crossdimensional/

Abstract

We propose a simple and straightforward way of creating powerful image representations via cross-dimensional weighting and aggregation of deep convolutional neural network layer outputs. We first present a generalized framework that encompasses a broad family of approaches and includes cross-dimensional pooling and weighting steps. We then propose specific non-parametric schemes for both spatial- and channel-wise weighting that boost the effect of highly active spatial responses and at the same time regulate burstiness effects. We experiment on different public datasets for image search and show that our approach outperforms the current state-of-the-art for approaches based on pre-trained networks. We also provide an easy-to-use, open source implementation that reproduces our results.

PDF ECCVW Semantic Scholar

Cite

Text

Kalantidis et al. "Cross-Dimensional Weighting for Aggregated Deep Convolutional Features." European Conference on Computer Vision Workshops, 2016. doi:10.1007/978-3-319-46604-0_48

Markdown

[Kalantidis et al. "Cross-Dimensional Weighting for Aggregated Deep Convolutional Features." European Conference on Computer Vision Workshops, 2016.](https://mlanthology.org/eccvw/2016/kalantidis2016eccvw-crossdimensional/) doi:10.1007/978-3-319-46604-0_48

BibTeX

@inproceedings{kalantidis2016eccvw-crossdimensional,
  title     = {{Cross-Dimensional Weighting for Aggregated Deep Convolutional Features}},
  author    = {Kalantidis, Yannis and Mellina, Clayton and Osindero, Simon},
  booktitle = {European Conference on Computer Vision Workshops},
  year      = {2016},
  pages     = {685-701},
  doi       = {10.1007/978-3-319-46604-0_48},
  url       = {https://mlanthology.org/eccvw/2016/kalantidis2016eccvw-crossdimensional/}
}