The truth about cats and dogs

Conference item

Template-based object detectors such as the deformable parts model of Felzenszwalb et al. [11] achieve state-of-the-art performance for a variety of object categories, but are still outperformed by simpler bag-of-words models for highly flexible objects such as cats and dogs. In these cases we propose to use the template-based model to detect a distinctive part for the class, followed by detecting the rest of the object via segmentation on image specific information learnt from that part. This approach is motivated by two ob- servations: (i) many object classes contain distinctive parts that can be detected very reliably by template-based detec- tors, whilst the entire object cannot; (ii) many classes (e.g. animals) have fairly homogeneous coloring and texture that can be used to segment the object once a sample is provided in an image. We show quantitatively that our method substantially outperforms whole-body template-based detectors for these highly deformable object categories, and indeed achieves accuracy comparable to the state-of-the-art on the PASCAL VOC competition, which includes other models such as bag-of-words. © 2011 IEEE.

Authors: Parkhi, O; Vedaldi, A; Jawahar, C; Zisserman, A

• Publication status: Published
• Peer review status: Peer reviewed
• Publisher: IEEE
• Host title: 2011 International Conference on Computer Vision
• Pages: 1427-1434
• Publication date: 2012-01-12
• Event title: International Conference on Computer Vision (ICCV 2011)
• Event location: Barcelona, Spain
• Event start date: 2011-11-06
• Event end date: 2011-11-13
• DOI: 10.1109/ICCV.2011.6126398
• EISSN: 2380-7504
• ISSN: 1550-5499
• EISBN: 978-1-4577-1102-2
• ISBN: 978-1-4577-1101-5
• Language: English
• Keywords: detectors; SBTMR; head; training; image color analysis; image segmentation; cats; image edge detection