Understanding Objects in Detail with Fine-grained Attributes

Andrea Vedaldi 1 Siddarth Mahendran 2 Stavros Tsogkas 3, 4 Subhransu Maji 5 Ross B. Girshick 6 Juho Kannala 7 Esa Rahtu 8 Iasonas Kokkinos 3, 4 Matthew B. Blaschko 3, 4 David Weiss 9 Ben Taskar 10 Karen Simonyan 1 Naomi Saphra 11 Sammy Mohamed 12
1 Department of Engineering Science
2 CIS - Center for Imaging Science
3 GALEN - Organ Modeling through Extraction, Representation and Understanding of Medical Image Content
Inria Saclay - Ile de France, Ecole Centrale Paris
4 CVN - Centre de vision numérique
5 Toyota Technological Institute
6 Department of Computer Science, University of Chicago
7 CMV - Center for Machine Vision Research
8 MVG - Machine Vision Group
9 Google Inc.
10 CSE - Department of Computer Science & Engineering
11 JHU - Johns Hopkins University
12 SBU - Stony Brook University [The State University of New York]
Abstract : We study the problem of understanding objects in detail, intended as recognizing a wide array of fine-grained object attributes. To this end, we introduce a dataset of 7,413 airplanes annotated in detail with parts and their attributes, leveraging images donated by airplane spotters and crowd-sourcing both the design and collection of the detailed annotations. We provide a number of insights that should help researchers interested in designing fine-grained datasets for other basic level categories. We show that the collected data can be used to study the relation between part detection and attribute prediction by diagnosing the performance of classifiers that pool information from different parts of an object. We note that the prediction of certain attributes can benefit substantially from accurate part detection. We also show that, differently from previous results in object detection, employing a large number of part templates can improve detection accuracy at the expenses of detection speed. We finally propose a coarse-to-fine approach to speed up detection through a hierarchical cascade algorithm.
Document type :
Conference papers
Liste complète des métadonnées

Cited literature [44 references]  Display  Hide  Download
https://hal.inria.fr/hal-00981125
Contributor : Matthew Blaschko <>
Submitted on : Monday, April 21, 2014 - 2:53:38 AM
Last modification on : Thursday, February 7, 2019 - 5:29:11 PM
Document(s) archivé(s) le : Monday, April 10, 2017 - 3:58:15 PM

File

OIDcvpr.pdf
Files produced by the author(s)

Identifiers

Collections

INRIA | EC-PARIS | CVN

Citation

Andrea Vedaldi, Siddarth Mahendran, Stavros Tsogkas, Subhransu Maji, Ross B. Girshick, et al.. Understanding Objects in Detail with Fine-grained Attributes. IEEE Conference on Computer Vision and Pattern Recognition, Jun 2014, Columbus, United States. ⟨10.1109/CVPR.2014.463⟩. ⟨hal-00981125⟩

Export

BibTeX TEI DC DCterms EndNote

Share

Metrics

Record views

603

Files downloads

627

Contact


archive-ouverte@inria.fr