NCJ Number
254594
Date Published
2019
Length
9 pages
Annotation
For convolutional neural network models that optimize an image embedding, this article proposes a method to highlight the regions of images that contribute most to pairwise similarity.
Abstract
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. This approach is generalized to embedding networks that use different pooling strategies and provides a simple mechanism to support image similarity searches on objects or sub-regions in the query image. (publisher abstract modified)
Date Published: January 1, 2019
Downloads
Similar Publications
- Lay understanding of forensic statistics: Evaluation of random match probabilities, likelihood ratios, and verbal equivalents.
- Do Observer Effects Matter? A Comment on Langenburg, Bochet, and Ford
- Testing Reliability of the Computational Age-At-Death Estimation Methods between Five Observers Using Three-Dimensional Image Data of the Pubic Symphysis