Abstract
The biologically inspired hierarchical model for object recognition, Hierarchical Model and X (HMAX), has attracted considerable attention in recent years. HMAX is robust (i.e. shift- and scale-invariant), but its use of random-patch-selection makes it sensitive to rotational deformation, which heavily limits its performance in object recognition. The main reason is that numerous randomly chosen patches are often orientation selective, thereby leading to mismatch. To address this issue, the authors propose a novel patch selection method for HMAX called saliency and keypoint-based patch selection (SKPS), which is based on a saliency (attention) mechanism and multi-scale keypoints. In contrast to the conventional random-patch-selection-based HMAX model that involves huge amounts of redundant information in feature extraction, the SKPS-based HMAX model (S-HMAX) extracts a very few features while offering promising distinctiveness. To show the effectiveness of S-HMAX, the authors apply it to object categorisation and conduct experiments on the CalTech101, TU Darmstadt, ImageNet and GRAZ01 databases. The experimental results demonstrate that S-HMAX outperforms conventional HMAX and is very comparable with existing architectures that have a similar framework.
| Original language | English |
|---|---|
| Pages (from-to) | 663-672 |
| Number of pages | 10 |
| Journal | IET Computer Vision |
| Volume | 9 |
| Issue number | 5 |
| DOIs | |
| Publication status | Published - 2015 Oct 1 |
Fingerprint
ASJC Scopus subject areas
- Computer Vision and Pattern Recognition
- Software
Cite this
Enhanced hierarchical model of object recognition based on a novel patch selection method in salient regions. / Lu, Yan Feng; Kang, Tae Koo; Zhang, Hua Zhen; Lim, Myo Taeg.
In: IET Computer Vision, Vol. 9, No. 5, 01.10.2015, p. 663-672.Research output: Contribution to journal › Article
}
TY - JOUR
T1 - Enhanced hierarchical model of object recognition based on a novel patch selection method in salient regions
AU - Lu, Yan Feng
AU - Kang, Tae Koo
AU - Zhang, Hua Zhen
AU - Lim, Myo Taeg
PY - 2015/10/1
Y1 - 2015/10/1
N2 - The biologically inspired hierarchical model for object recognition, Hierarchical Model and X (HMAX), has attracted considerable attention in recent years. HMAX is robust (i.e. shift- and scale-invariant), but its use of random-patch-selection makes it sensitive to rotational deformation, which heavily limits its performance in object recognition. The main reason is that numerous randomly chosen patches are often orientation selective, thereby leading to mismatch. To address this issue, the authors propose a novel patch selection method for HMAX called saliency and keypoint-based patch selection (SKPS), which is based on a saliency (attention) mechanism and multi-scale keypoints. In contrast to the conventional random-patch-selection-based HMAX model that involves huge amounts of redundant information in feature extraction, the SKPS-based HMAX model (S-HMAX) extracts a very few features while offering promising distinctiveness. To show the effectiveness of S-HMAX, the authors apply it to object categorisation and conduct experiments on the CalTech101, TU Darmstadt, ImageNet and GRAZ01 databases. The experimental results demonstrate that S-HMAX outperforms conventional HMAX and is very comparable with existing architectures that have a similar framework.
AB - The biologically inspired hierarchical model for object recognition, Hierarchical Model and X (HMAX), has attracted considerable attention in recent years. HMAX is robust (i.e. shift- and scale-invariant), but its use of random-patch-selection makes it sensitive to rotational deformation, which heavily limits its performance in object recognition. The main reason is that numerous randomly chosen patches are often orientation selective, thereby leading to mismatch. To address this issue, the authors propose a novel patch selection method for HMAX called saliency and keypoint-based patch selection (SKPS), which is based on a saliency (attention) mechanism and multi-scale keypoints. In contrast to the conventional random-patch-selection-based HMAX model that involves huge amounts of redundant information in feature extraction, the SKPS-based HMAX model (S-HMAX) extracts a very few features while offering promising distinctiveness. To show the effectiveness of S-HMAX, the authors apply it to object categorisation and conduct experiments on the CalTech101, TU Darmstadt, ImageNet and GRAZ01 databases. The experimental results demonstrate that S-HMAX outperforms conventional HMAX and is very comparable with existing architectures that have a similar framework.
UR - http://www.scopus.com/inward/record.url?scp=84942028366&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84942028366&partnerID=8YFLogxK
U2 - 10.1049/iet-cvi.2014.0249
DO - 10.1049/iet-cvi.2014.0249
M3 - Article
AN - SCOPUS:84942028366
VL - 9
SP - 663
EP - 672
JO - IET Computer Vision
JF - IET Computer Vision
SN - 1751-9632
IS - 5
ER -