ContextLocNet: Context-Aware Deep Network Models for Weakly Supervised Localization

Vadim Kantorov 1, 2 Maxime Oquab 3, 2, 1 Minsu Cho 2, 1 Ivan Laptev 1, 2
1 WILLOW - Models of visual object recognition and scene understanding
DI-ENS - Département d'informatique de l'École normale supérieure, Inria Paris-Rocquencourt, CNRS - Centre National de la Recherche Scientifique : UMR8548
Abstract : We aim to localize objects in images using image-level supervision only. Previous approaches to this problem mainly focus on discrim-inative object regions and often fail to locate precise object boundaries. We address this problem by introducing two types of context-aware guidance models, additive and contrastive models, that leverage their surrounding context regions to improve localization. The additive model encourages the predicted object region to be supported by its surrounding context region. The contrastive model encourages the predicted object region to be outstanding from its surrounding context region. Our approach benefits from the recent success of convolutional neural networks for object recognition and extends Fast R-CNN to weakly supervised object localization. Extensive experimental evaluation on the PASCAL VOC 2007 and 2012 benchmarks shows that our context-aware approach significantly improves weakly supervised localization and detection.
Document type :
Conference papers
Complete list of metadatas

Cited literature [40 references]  Display  Hide  Download

https://hal.inria.fr/hal-01421772
Contributor : Vadim Kantorov <>
Submitted on : Thursday, December 22, 2016 - 6:25:57 PM
Last modification on : Thursday, February 7, 2019 - 3:49:20 PM
Long-term archiving on: Thursday, March 23, 2017 - 12:12:19 PM

File

contextlocnet_eccv2016.pdf
Files produced by the author(s)

Identifiers

Collections

Citation

Vadim Kantorov, Maxime Oquab, Minsu Cho, Ivan Laptev. ContextLocNet: Context-Aware Deep Network Models for Weakly Supervised Localization. ECCV 2016, Oct 2016, Amsterdam, Netherlands. pp.350 - 365, ⟨10.1007/978-3-319-46454-1_22⟩. ⟨hal-01421772⟩

Share

Metrics

Record views

466

Files downloads

697