Semantic Apparatus – Unsupervised semantic discovery through visual patterns detection

Cited by Lee Sonogan

Using a Semantic Layer to Propel a Data-Driven Culture - RTInsights

Abstract by Francesco Pelosin, Andrea Gasparetto, Andrea Albarelli, Andrea Torsello

We propose a new fast fully unsupervised method to discover semantic patterns. Our algorithm is able to hierarchically find visual categories and produce a segmentation mask where previous methods fail. Through the modeling of what is a visual pattern in an image, we introduce the notion of “semantic levels” and devise a conceptual framework along with measures and a dedicated benchmark dataset for future comparisons. Our algorithm is composed by two phases. A filtering phase, which selects semantical hotsposts by means of an accumulator space, then a clustering phase which propagates the semantic properties of the hotspots on a superpixels basis. We provide both qualitative and quantitative experimental validation, achieving optimal results in terms of robustness to noise and semantic consistency. We also made code and dataset publicly available.

Publication: Cornell University (Peer-Reviewed Journal)

Pub Date: 24 Feb, 2021 Doi:

Keywords: Computer Vision, Pattern Recognition, Semantic Discovery, Visual Patterns, Unsupervised Detection (Plenty more sections and references in this research article)

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.