Sequential auxiliary particle belief propagation
This paper discloses a novel algorithm for efficient inference in undirected graphical models using Sequential Monte Carlo (SMC) based numerical approximation techniques. The methodology developed, titled "Auxiliary Particle Belief Propagation", extends the applicability of the much celebr...
| Main Authors: | , , , |
|---|---|
| Format: | Journal article |
| Language: | English |
| Published: |
2005
|
| _version_ | 1797105738149003264 |
|---|---|
| author | Briers, M Doucet, A Singh, S IEEE |
| author_facet | Briers, M Doucet, A Singh, S IEEE |
| author_sort | Briers, M |
| collection | OXFORD |
| description | This paper discloses a novel algorithm for efficient inference in undirected graphical models using Sequential Monte Carlo (SMC) based numerical approximation techniques. The methodology developed, titled "Auxiliary Particle Belief Propagation", extends the applicability of the much celebrated (Loopy) Belief Propagation (LBP) algorithm to non-linear, non-Gaussian models, whilst retaining a computational cost that is linear in the number of sample points (or particles). Furthermore, we provide an additional extension to this technique by analysing temporally evolving graphical models, a problem which remains largely unexplored in the scientific literature. The work presented is thus a general framework that can be applied to a plethora of novel distributed fusion problems. In this paper, we apply our inference algorithm to the (sequential problem of) articulated object tracking. © 2005 IEEE. |
| first_indexed | 2024-03-07T06:51:38Z |
| format | Journal article |
| id | oxford-uuid:fcc2473b-1967-4e67-8aa1-e763a7c605dd |
| institution | University of Oxford |
| language | English |
| last_indexed | 2024-03-07T06:51:38Z |
| publishDate | 2005 |
| record_format | dspace |
| spelling | oxford-uuid:fcc2473b-1967-4e67-8aa1-e763a7c605dd2022-03-27T13:23:20ZSequential auxiliary particle belief propagationJournal articlehttp://purl.org/coar/resource_type/c_dcae04bcuuid:fcc2473b-1967-4e67-8aa1-e763a7c605ddEnglishSymplectic Elements at Oxford2005Briers, MDoucet, ASingh, SIEEEThis paper discloses a novel algorithm for efficient inference in undirected graphical models using Sequential Monte Carlo (SMC) based numerical approximation techniques. The methodology developed, titled "Auxiliary Particle Belief Propagation", extends the applicability of the much celebrated (Loopy) Belief Propagation (LBP) algorithm to non-linear, non-Gaussian models, whilst retaining a computational cost that is linear in the number of sample points (or particles). Furthermore, we provide an additional extension to this technique by analysing temporally evolving graphical models, a problem which remains largely unexplored in the scientific literature. The work presented is thus a general framework that can be applied to a plethora of novel distributed fusion problems. In this paper, we apply our inference algorithm to the (sequential problem of) articulated object tracking. © 2005 IEEE. |
| spellingShingle | Briers, M Doucet, A Singh, S IEEE Sequential auxiliary particle belief propagation |
| title | Sequential auxiliary particle belief propagation |
| title_full | Sequential auxiliary particle belief propagation |
| title_fullStr | Sequential auxiliary particle belief propagation |
| title_full_unstemmed | Sequential auxiliary particle belief propagation |
| title_short | Sequential auxiliary particle belief propagation |
| title_sort | sequential auxiliary particle belief propagation |
| work_keys_str_mv | AT briersm sequentialauxiliaryparticlebeliefpropagation AT douceta sequentialauxiliaryparticlebeliefpropagation AT singhs sequentialauxiliaryparticlebeliefpropagation AT ieee sequentialauxiliaryparticlebeliefpropagation |