Natural Scene Classification Based on Integrated Topic Simplex

Tang YINGJUN; Xu DE; Yang XU; Liu QIFANG

doi:10.1587/transinf.E92.D.1811

Natural Scene Classification Based on Integrated Topic Simplex

Tang YINGJUN, Xu DE, Yang XU, Liu QIFANG

Full Text Views

0

Cite this

Summary :

We present a novel model named Integrated Latent Topic Model (ILTM), to learn and recognize natural scene category. Unlike previous work, which considered the discrepancy and common property separately among all categories, Our approach combines universal topics from all categories with specific topics from each category. As a result, the model is implemented to produce a few but specific topics and more generic topics among categories, and each category is represented in a different topics simplex, which correlates well with human scene understanding. We investigate the classification performance with variable scene category tasks. The experiments have shown our model outperforms latent-space methods with less training data.

Publication: IEICE TRANSACTIONS on Information Vol.E92-D No.9 pp.1811-1814

Publication Date: 2009/09/01

Publicized

Online ISSN: 1745-1361

DOI: 10.1587/transinf.E92.D.1811

Type of Manuscript: LETTER

Category: Image Recognition, Computer Vision

Cite this

Copy

Tang YINGJUN, Xu DE, Yang XU, Liu QIFANG, "Natural Scene Classification Based on Integrated Topic Simplex" in IEICE TRANSACTIONS on Information, vol. E92-D, no. 9, pp. 1811-1814, September 2009, doi: 10.1587/transinf.E92.D.1811.
Abstract: We present a novel model named Integrated Latent Topic Model (ILTM), to learn and recognize natural scene category. Unlike previous work, which considered the discrepancy and common property separately among all categories, Our approach combines universal topics from all categories with specific topics from each category. As a result, the model is implemented to produce a few but specific topics and more generic topics among categories, and each category is represented in a different topics simplex, which correlates well with human scene understanding. We investigate the classification performance with variable scene category tasks. The experiments have shown our model outperforms latent-space methods with less training data.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.E92.D.1811/_p

Copy

@ARTICLE{e92-d_9_1811,
author={Tang YINGJUN, Xu DE, Yang XU, Liu QIFANG, },
journal={IEICE TRANSACTIONS on Information},
title={Natural Scene Classification Based on Integrated Topic Simplex},
year={2009},
volume={E92-D},
number={9},
pages={1811-1814},
abstract={We present a novel model named Integrated Latent Topic Model (ILTM), to learn and recognize natural scene category. Unlike previous work, which considered the discrepancy and common property separately among all categories, Our approach combines universal topics from all categories with specific topics from each category. As a result, the model is implemented to produce a few but specific topics and more generic topics among categories, and each category is represented in a different topics simplex, which correlates well with human scene understanding. We investigate the classification performance with variable scene category tasks. The experiments have shown our model outperforms latent-space methods with less training data.},
keywords={},
doi={10.1587/transinf.E92.D.1811},
ISSN={1745-1361},
month={September},}

Copy

TY - JOUR
TI - Natural Scene Classification Based on Integrated Topic Simplex
T2 - IEICE TRANSACTIONS on Information
SP - 1811
EP - 1814
AU - Tang YINGJUN
AU - Xu DE
AU - Yang XU
AU - Liu QIFANG
PY - 2009
DO - 10.1587/transinf.E92.D.1811
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E92-D
IS - 9
JA - IEICE TRANSACTIONS on Information
Y1 - September 2009
AB - We present a novel model named Integrated Latent Topic Model (ILTM), to learn and recognize natural scene category. Unlike previous work, which considered the discrepancy and common property separately among all categories, Our approach combines universal topics from all categories with specific topics from each category. As a result, the model is implemented to produce a few but specific topics and more generic topics among categories, and each category is represented in a different topics simplex, which correlates well with human scene understanding. We investigate the classification performance with variable scene category tasks. The experiments have shown our model outperforms latent-space methods with less training data.
ER -