Image Indexing and Retrieval with Pachinko Allocation Model: Application on Local and Global Features

Boulemden, Ahmed; Tlili, Yamina

doi:10.1007/978-3-642-32541-0_12

Cited by 3 publications

(3 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It does so by extending the concept of latent topics to be distributions not only over the visual words but also over other latent topics. Using image data from large-scale databases, Boulemden and Tilli [4] reported improved performance of PAM-based latent topic representation in image retrieval operation.…”

Section: Related Workmentioning

confidence: 99%

“…Graphical model-based latent topic frameworks for image retrieval fall into two fundamental categories such as (i) directed topic models and (ii) undirected topic models. The former category involves models based on directed graphs and the most successful approaches toward this direction are Probabilistic Latent Semantic Analysis (PLSA) [1], Latent Dirichlet Allocation (LDA) [2], Correlated Topic Models (CTM) [3] and Pachinko Allocation Model (PAM) [4]. On the contrary, undirected topic modeling frameworks encode the joint distribution by means of undirected graphs.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A Hybrid Deep Learning Architecture for Latent Topic-based Image Retrieval

Arun

Govindan

2018

Data Sci. Eng.

View full text Add to dashboard Cite

Learning effective feature descriptors that bridge the semantic gap between low-level visual features directly extracted from image pixels and the corresponding high-level semantics perceived by humans is a challenging task in image retrieval. This paper proposes a hybrid deep learning architecture (HDLA) that generates sparse latent topic-based representation with the objective of minimizing the semantic gap problem in image retrieval. In fact, HDLA has a deep network structure with a constrained replicated Softmax Model in the lower layer and constrained restricted Boltzmann machines in the upper layers. The advantage of HDLA is that there exist nonnegativity restrictions on the model weights together with ' 1-sparsity enforced over the activations of the hidden layer nodes of the network. This, in turn, enhances the modeling power of the network and leads to sparse, parts-based latent topic representation of images. Experimental results on various benchmark datasets show that the proposed model exhibits better generalization ability and the resulting highlevel abstraction yields better retrieval performance as compared to state-of-the-art latent topic-based image representation schemes.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

A Hybrid Deep Learning Architecture for Latent Topic-based Image Retrieval

Arun

Govindan

2018

Data Sci. Eng.

View full text Add to dashboard Cite

show abstract

“…Ma et al [10] proposed a new generative model, the Labelled Four-Level Pachinko Allocation Model (L-F-L-PAM), to infer training data and capture correlations among multiple labels. Boulemden et al [11] used PAM with local image features extracted by scale-invariant feature transform technique in a content-image retrieval task. Bakalov et al [12] proposed the Labelled Pachinko Allocation Model and Labelled Pachinko Allocation-List Model.…”

Section: Related Workmentioning

confidence: 99%

A PAM-based ontology concept and hierarchy learning method

Wang

Zhao

2013

Journal of Information Science

View full text Add to dashboard Cite

As a shared conceptual model that can express knowledge and a modelling tool that can describe conceptual model in the semantic and knowledge level, ontology plays an important role in the related fields of the Semantic Web, natural language processing and information retrieval. Ontology learning is a series of methods and technologies to construct ontology automatically or semi-automatically. Concept and hierarchy learning are the most important parts of the ontology construction. This paper proposes an ontology concept and hierarchy learning method based on the Pachinko Allocation Model. The above problem is transformed into a probability and statistical inference problem by building an ontology concept learning model. Gibbs sampling is used to estimate the parameters. Then, using the ontology concept generation algorithm based on WordNet, an abstract description of the ontology concept is obtained. Experimental results on the standard test dataset show that the proposed method can offer an effective solution to ontology concept and hierarchy learning.

show abstract