Segmentation-Aware Deformable Part Models

Abstract : In this work we propose a technique to combine bottom-up segmentation, coming in the form of SLIC superpixels, with sliding window detectors, such as Deformable Part Models (DPMs). The merit of our approach lies in 'cleaning up' the low-level HOG features by exploiting the spatial support of SLIC superpixels; this can be understood as using segmentation to split the feature variation into object-specific and background changes. Rather than committing to a single segmentation we use a large pool of SLIC superpixels and combine them in a scale-, position-and object-dependent manner to build soft segmentation masks. The segmentation masks can be computed fast enough to repeat this process over every candidate window, during training and detection, for both the root and part filters of DPMs. We use these masks to construct enhanced, background-invariant features to train DPMs. We test our approach on the PASCAL VOC 2007, outperforming the standard DPM in 17 out of 20 classes, yielding an average increase of 1.7% AP. Additionally, we demonstrate the robustness of this approach, extending it to dense SIFT descriptors for large displacement optical flow.
Type de document :
Communication dans un congrès
IEEE Conference in Computer Vision and Pattern Recognition, Jun 2014, Greater Columbus Convention Center in Columbus, Ohio., United States. pp.168 - 175, 2014, 〈http://www.pamitc.org/cvpr14/〉. 〈10.1109/CVPR.2014.29〉
Liste complète des métadonnées

Littérature citée [42 références]  Voir  Masquer  Télécharger

https://hal.archives-ouvertes.fr/hal-01109286
Contributeur : Stavros Tsogkas <>
Soumis le : lundi 2 février 2015 - 14:41:01
Dernière modification le : jeudi 7 février 2019 - 17:29:35
Document(s) archivé(s) le : mercredi 27 mai 2015 - 13:40:49

Fichier

Eduard_SegmentationAware_CVPR1...
Fichiers produits par l'(les) auteur(s)

Identifiants

Collections

Citation

Eduard Trulls, Stavros Tsogkas, Iasonas Kokkinos, Alberto Sanfeliu, Francesc Moreno-Noguer. Segmentation-Aware Deformable Part Models. IEEE Conference in Computer Vision and Pattern Recognition, Jun 2014, Greater Columbus Convention Center in Columbus, Ohio., United States. pp.168 - 175, 2014, 〈http://www.pamitc.org/cvpr14/〉. 〈10.1109/CVPR.2014.29〉. 〈hal-01109286〉

Partager

Métriques

Consultations de la notice

253

Téléchargements de fichiers

227