Prototypicality Effects in Global Semantic Description of Objects

Abstract

In this paper, we introduce a novel approach for semantic description of object features based on the prototypicality effects of the Prototype Theory. Our prototype-based description model encodes and stores the semantic meaning of an object, while describing its features using the semantic prototype computed by CNN-classifications models. Our method uses semantic prototypes to create discriminative descriptor signatures that describe an object highlighting its most distinctive features within the category. Our experiments show that: i) our descriptor preserves the semantic information used by the CNN-models in classification tasks; ii) our distance metric can be used as the object's typicality score; iii) our descriptor signatures are semantically interpretable and enables the simulation of the prototypical organization of objects within a category.

Cite

Text

Pino et al. "Prototypicality Effects in Global Semantic Description of Objects." IEEE/CVF Winter Conference on Applications of Computer Vision, 2019. doi:10.1109/WACV.2019.00136

Markdown

[Pino et al. "Prototypicality Effects in Global Semantic Description of Objects." IEEE/CVF Winter Conference on Applications of Computer Vision, 2019.](https://mlanthology.org/wacv/2019/pino2019wacv-prototypicality/) doi:10.1109/WACV.2019.00136

BibTeX

@inproceedings{pino2019wacv-prototypicality,
  title     = {{Prototypicality Effects in Global Semantic Description of Objects}},
  author    = {Pino, Omar Vidal and Nascimento, Erickson R. and Campos, Mario F. M.},
  booktitle = {IEEE/CVF Winter Conference on Applications of Computer Vision},
  year      = {2019},
  pages     = {1233-1242},
  doi       = {10.1109/WACV.2019.00136},
  url       = {https://mlanthology.org/wacv/2019/pino2019wacv-prototypicality/}
}