Online learning of mixture experts for real‐time tracking
Template tracking has been extensively investigated in computer vision to track objects for various applications. Tracking based on gradient descent algorithm using image gradient is one of the most popular object tracking method. However, it is difficult to define the relationship between the obser...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2016-09-01
|
| Series: | IET Computer Vision |
| Subjects: | |
| Online Access: | https://doi.org/10.1049/iet-cvi.2015.0210 |
| _version_ | 1797684962212708352 |
|---|---|
| author | S. Gu Z. Ma M. Xie Z. Chen |
| author_facet | S. Gu Z. Ma M. Xie Z. Chen |
| author_sort | S. Gu |
| collection | DOAJ |
| description | Template tracking has been extensively investigated in computer vision to track objects for various applications. Tracking based on gradient descent algorithm using image gradient is one of the most popular object tracking method. However, it is difficult to define the relationship between the observed data set and the warping function due to the unobserved heterogeneity of the data set which inevitably results in poor tracking performance. This study proposes a novel method based on hierarchical mixture of expert to perform robust, real‐time tracking from stationary cameras. By extending the idea of hyperplane approximation, the proposed approach establishes a hierarchical mixture of generalised linear regression model instead of a single model which reduces the non‐linear error. The experiments’ results show significant improvement over the traditional hyperplane approximation (HA) approach. |
| first_indexed | 2024-03-12T00:37:26Z |
| format | Article |
| id | doaj.art-b44bbff96656440a9b3c23a0928e23a0 |
| institution | Directory Open Access Journal |
| issn | 1751-9632 1751-9640 |
| language | English |
| last_indexed | 2024-03-12T00:37:26Z |
| publishDate | 2016-09-01 |
| publisher | Wiley |
| record_format | Article |
| series | IET Computer Vision |
| spelling | doaj.art-b44bbff96656440a9b3c23a0928e23a02023-09-15T09:26:26ZengWileyIET Computer Vision1751-96321751-96402016-09-0110658559210.1049/iet-cvi.2015.0210Online learning of mixture experts for real‐time trackingS. Gu0Z. Ma1M. Xie2Z. Chen3School of Communication and Information EngineeringUniversity of Electronic Science and Technology of ChinaChengduPeople's Republic of ChinaSchool of Communication and Information EngineeringUniversity of Electronic Science and Technology of ChinaChengduPeople's Republic of ChinaSchool of Electronic EngineeringUniversity of Electronic Science and Technology of ChinaChengduPeople's Republic of ChinaFaculty of Computer ScienceMemorial University of NewfoundlandSt. John'sCanadaTemplate tracking has been extensively investigated in computer vision to track objects for various applications. Tracking based on gradient descent algorithm using image gradient is one of the most popular object tracking method. However, it is difficult to define the relationship between the observed data set and the warping function due to the unobserved heterogeneity of the data set which inevitably results in poor tracking performance. This study proposes a novel method based on hierarchical mixture of expert to perform robust, real‐time tracking from stationary cameras. By extending the idea of hyperplane approximation, the proposed approach establishes a hierarchical mixture of generalised linear regression model instead of a single model which reduces the non‐linear error. The experiments’ results show significant improvement over the traditional hyperplane approximation (HA) approach.https://doi.org/10.1049/iet-cvi.2015.0210online mixture experts learningtemplate trackingcomputer visiongradient descent algorithmobject tracking methodobserved data set |
| spellingShingle | S. Gu Z. Ma M. Xie Z. Chen Online learning of mixture experts for real‐time tracking IET Computer Vision online mixture experts learning template tracking computer vision gradient descent algorithm object tracking method observed data set |
| title | Online learning of mixture experts for real‐time tracking |
| title_full | Online learning of mixture experts for real‐time tracking |
| title_fullStr | Online learning of mixture experts for real‐time tracking |
| title_full_unstemmed | Online learning of mixture experts for real‐time tracking |
| title_short | Online learning of mixture experts for real‐time tracking |
| title_sort | online learning of mixture experts for real time tracking |
| topic | online mixture experts learning template tracking computer vision gradient descent algorithm object tracking method observed data set |
| url | https://doi.org/10.1049/iet-cvi.2015.0210 |
| work_keys_str_mv | AT sgu onlinelearningofmixtureexpertsforrealtimetracking AT zma onlinelearningofmixtureexpertsforrealtimetracking AT mxie onlinelearningofmixtureexpertsforrealtimetracking AT zchen onlinelearningofmixtureexpertsforrealtimetracking |