Interpretable Visual Models for Human Perception-Based Object Retrieval - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Conference Papers Year : 2011

Interpretable Visual Models for Human Perception-Based Object Retrieval


Understanding the results returned by automatic visual concept detectors is often a tricky task making users uncomfortable with these technologies. In this paper we attempt to build humanly interpretable visual models, allowing the user to visually understand the underlying semantic. We therefore propose a supervised multiple instance learning algorithm that selects as few as possible discriminant local features for a given object category. The method finds its roots in the lasso theory where a L1-regularization term is introduced in order to constraint the loss function, and subsequently produce sparser solutions. Efficient resolution of the lasso path is achieved through a boosting-like procedure inspired by BLasso algorithm. Quantitatively, our method achieves similar performance as current state-of-the-art, and qualitatively, it allows users to construct their own model from the original set of patches learned, thus allowing for more compound semantic queries.
Fichier principal
Vignette du fichier
ICMR-rebai.pdf (1.39 Mo) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-00642232 , version 1 (17-11-2011)



Ahmed-Riadh Rebai, Alexis Joly, Nozha Boujemaa. Interpretable Visual Models for Human Perception-Based Object Retrieval. ICMR'11 - First ACM International Conference on Multimedia Retrieval, Apr 2011, Trento, Italy. pp.21:1--21:8, ⟨10.1145/1991996.1992017⟩. ⟨hal-00642232⟩
230 View
241 Download



Gmail Facebook X LinkedIn More