Visualizing Deep Similarity Networks

Abby Stylianou, Richard Souvenir, Robert Pless

WACV 2019 pp. 2029-2037

doi:10.1109/WACV.2019.00220 /wacv/2019/stylianou2019wacv-visualizing/

Abstract

For convolutional neural network models that optimize an image embedding, we propose a method to highlight the regions of images that contribute most to pairwise similarity. This work is a corollary to the visualization tools developed for classification networks, but applicable to the problem domains better suited to similarity learning. The visualization shows how similarity networks that are fine-tuned learn to focus on different features. We also generalize our approach to embedding networks that use different pooling strategies and provide a simple mechanism to support image similarity searches on objects or sub-regions in the query image.

PDF WACV Semantic Scholar

Cite

Text

Stylianou et al. "Visualizing Deep Similarity Networks." IEEE/CVF Winter Conference on Applications of Computer Vision, 2019. doi:10.1109/WACV.2019.00220

Markdown

[Stylianou et al. "Visualizing Deep Similarity Networks." IEEE/CVF Winter Conference on Applications of Computer Vision, 2019.](https://mlanthology.org/wacv/2019/stylianou2019wacv-visualizing/) doi:10.1109/WACV.2019.00220

BibTeX

@inproceedings{stylianou2019wacv-visualizing,
  title     = {{Visualizing Deep Similarity Networks}},
  author    = {Stylianou, Abby and Souvenir, Richard and Pless, Robert},
  booktitle = {IEEE/CVF Winter Conference on Applications of Computer Vision},
  year      = {2019},
  pages     = {2029-2037},
  doi       = {10.1109/WACV.2019.00220},
  url       = {https://mlanthology.org/wacv/2019/stylianou2019wacv-visualizing/}
}