Semantic Image Collection Summarization with Frequent Subgraph Mining
With Andrea Pasini - Politecnico di Torino
Every day we generate large amounts of images and videos, including personal galleries, product catalogues and social data.
Image collection summarization, with the aim of displaying the important highlights with a few manageable pieces of information, is fundamental to understand and organize the content of these collections.
We propose SImS, an image collection summarization technique based on frequent subgraph mining.
Differently from previous methods, mainly based on low level visual features and image tags, our method exploits scene graphs to represent images.
Our results are interpretable and provide more powerful semantic information with respect to previous techniques, where the summary is a subset of the collection in terms of images or image patches.
Online on https://smartdata.polito.it/category/smarttalks/