Deformable Part-based Fully Convolutional Network for Object Detection

Abstract : Existing region-based object detectors are limited to regions with fixed box geometry to represent objects, even if those are highly non-rectangular. In this paper we introduce DP-FCN, a deep model for object detection which explicitly adapts to shapes of objects with deformable parts. Without additional annotations, it learns to focus on discriminative elements and to align them, and simultaneously brings more invariance for classification and geometric information to refine localization. DP-FCN is composed of three main modules: a Fully Convolutional Network to efficiently maintain spatial resolution, a deformable part-based RoI pooling layer to optimize positions of parts and build invariance, and a deformation-aware localization module explicitly exploiting displacements of parts to improve accuracy of bounding box regression. We experimentally validate our model and show significant gains. DP-FCN achieves state-of-the-art performances of 83.1% and 80.9% on PASCAL VOC 2007 and 2012 with VOC data only.
Liste complète des métadonnées

https://hal.archives-ouvertes.fr/hal-01637070
Contributor : Taylor Mordan <>
Submitted on : Friday, November 17, 2017 - 11:47:00 AM
Last modification on : Thursday, March 21, 2019 - 2:44:12 PM

Links full text

Identifiers

  • HAL Id : hal-01637070, version 1
  • ARXIV : 1707.06175

Citation

Taylor Mordan, Nicolas Thome, Matthieu Cord, Gilles Henaff. Deformable Part-based Fully Convolutional Network for Object Detection. British Machine Vision Conference (BMVC), Sep 2017, London, United Kingdom. ⟨hal-01637070⟩

Share

Metrics

Record views

119