A Non-Structural Representation Scheme for Articulated Shapes

Creative Commons License

Gençtav A., TARI Z. S.

JOURNAL OF IMAGING, vol.4, no.10, 2018 (ESCI) identifier identifier

  • Publication Type: Article / Article
  • Volume: 4 Issue: 10
  • Publication Date: 2018
  • Doi Number: 10.3390/jimaging4100115
  • Journal Name: JOURNAL OF IMAGING
  • Journal Indexes: Emerging Sources Citation Index (ESCI), Scopus
  • Keywords: 2D shapes, shape representation, Robust Principal Component Analysis (RPCA), t-Distributed Stochastic Neighbor Embedding (t-SNE), DISTANCE, CONTEXT
  • TED University Affiliated: No


Articulated shapes are successfully represented by structural representations which are organized in the form of graphs of shape components. We present an alternative representation scheme which is equally powerful but does not require explicit modeling or discovery of structural relations. The key element in our scheme is a novel multi scale pixel-based distinctness measure which implicitly quantifies how rare a particular pixel is in terms of its geometry with respect to all pixels of the shape. The spatial distribution of the distinctness yields a partitioning of the shape into a set of regions. The proposed representation is a collection of size normalized probability distribution of the distinctness over regions over shape dependent scales. We test the proposed representation on a clustering task.