Patch-level spatial layout for classification and weakly supervised localization

Valentina Zadrija 1 Josip Krapac 1 Jakob Verbeek 2 Siniša Šegvić 1
1 FER - Faculty of Electrical Engineering and Computing [Zagreb]
2 LEAR - Learning and recognition in vision
Inria Grenoble - Rhône-Alpes, LJK - Laboratoire Jean Kuntzmann, INPG - Institut National Polytechnique de Grenoble
Abstract : We propose a discriminative patch-level spatial layout model suitable for training with weak supervision. We start from a block-sparse model of patch appearance based on the normalized Fisher vector representation. The appearance model is responsible for i) selecting a discriminative subset of visual words, and ii) identifying distinctive patches assigned to the selected subset. These patches are further filtered by a sparse spatial model operating on a novel representation of pairwise patch layout. We have evaluated the proposed pipeline in image classification and weakly supervised localization experiments on a public traffic sign dataset. The results show significant advantage of the proposed spatial model over state of the art appearance models.
Document type :
Conference papers
Complete list of metadatas

Cited literature [38 references]  Display  Hide  Download
https://hal.inria.fr/hal-01186677
Contributor : Thoth Team <>
Submitted on : Thursday, August 27, 2015 - 3:12:25 PM
Last modification on : Monday, December 17, 2018 - 11:22:02 AM
Long-term archiving on : Saturday, November 28, 2015 - 10:19:52 AM

File

main68.pdf
Files produced by the author(s)

Identifiers

Collections

UGA | INRIA | LJK | LJK_GI | LJK_GI_LEAR

Citation

Valentina Zadrija, Josip Krapac, Jakob Verbeek, Siniša Šegvić. Patch-level spatial layout for classification and weakly supervised localization. GCPR - 37th German Conference on Pattern Recognition, Oct 2015, Aachen, Germany. pp.492-503, ⟨10.1007/978-3-319-24947-6_41⟩. ⟨hal-01186677⟩

Export

BibTeX TEI DC DCterms EndNote

Share

Metrics

Record views

494

Files downloads

455

Contact


archive-ouverte@inria.fr