High order pLSA for indexing tagged images

Nikolopoulos, Spiros; Zafeiriou, Stefanos; Patras, Ioannis; Kompatsiaris, Ioannis

doi:10.1016/j.sigpro.2012.08.004

Cited by 13 publications

(9 citation statements)

References 52 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Text categorisation [19], vocabulary reduction [20], visual encoding [21], image recognition [22] or even video retrieval [23] are some of the applications where topic models have been successfully used.…”

Section: Current Limitations and Trendsmentioning

confidence: 99%

Sparse multi-modal probabilistic latent semantic analysis for single-image super-resolution

Fernández-Beltrán

Pla

2018

Signal Processing

View full text Add to dashboard Cite

This paper presents a novel single-image super-resolution (SR) approach based on latent topics in order to take advantage of the semantics pervading the topic space when super-resolving images. Image semantics has shown to be useful to relieve the ill-posed nature of the SR problem, however the most accepted clustering-based approach used to define semantic concepts limits the capability of representing complex visual relationships. The proposed approach provides a new probabilistic perspective where the SR process is performed according to the semantics encapsulated by a new topic model, the Sparse Multimodal probabilistic Latent Semantic Analysis (sMpLSA). Firstly, the sMpLSA model is formulated. Subsequently, a new SR framework based on sMpLSA is defined. Finally, an experimental comparison is conducted using seven learningbased SR methods over three different image datasets. Experiments reveal the potential of latent topics in SR by reporting that the proposed approach is able to provide a competitive performance.

show abstract

Section: Current Limitations and Trendsmentioning

confidence: 99%

Sparse multi-modal probabilistic latent semantic analysis for single-image super-resolution

Fernández-Beltrán

Pla

2018

Signal Processing

View full text Add to dashboard Cite

show abstract

“…Then, a Gaussian-multinomial pLSA (GM-pLSA) model [27] was presented to learn multimodal correlations from the image data by applying continuous feature vectors. Furthermore, the work in [28] extended pLSA to a higher-order formalism, so as to become applicable for more than two observable variables. However, pLSA-based models are incomplete in that they provide no probabilistic restriction on how to generate the training data.…”

Section: Topic Models For Image Annotationmentioning

confidence: 99%

Class-specific Gaussian-multinomial latent Dirichlet allocation for image annotation

Qian

Zhong

Wang

2015

EURASIP J. Adv. Signal Process.

View full text Add to dashboard Cite

Image annotation has been a challenging problem due to the well-known semantic gap between two heterogeneous information modalities, i.e., the visual modality referring to low-level visual features and the semantic modality referring to high-level human concepts. To bridge the semantic gap, we present an extension of latent Dirichlet allocation (LDA), denoted as class-specific Gaussian-multinomial latent Dirichlet allocation (csGM-LDA), in an effort to simulate the human's visual perception system. An analysis of previous supervised LDA models shows that the topics discovered by generative LDA models are driven by general image regularities rather than the semantic regularities for image annotation. To address this, csGM-LDA is introduced by using class supervision at the level of visual features for multimodal topic modeling. The csGM-LDA model combines the labeling strength of topic supervision with the flexibility of topic discovery, and the modeling problem can be effectively solved by a variational expectation-maximization (EM) algorithm. Moreover, as natural images usually generate an enormous size of high-dimensional data in annotation applications, an efficient descriptor based on Laplacian regularized uncorrelated tensor representation is proposed for explicitly exploiting the manifold structures in the high-order image space. Experimental results on two standard annotation datasets have shown the effectiveness of the proposed method by comparing with several state-of-the-art annotation methods.

show abstract

“…More specifically, with respect to the heterogeneity of the collected data we have incorporated in the proposed framework novel methods [7], [8] for jointly handling images and text, which are typically the most frequent resources contributed by users. In the first case [7], the rationale is to project the features extracted from the heterogeneous resources into a new space (e.g.…”

Section: Proposed Approachmentioning

confidence: 99%

“…In the first case [7], the rationale is to project the features extracted from the heterogeneous resources into a new space (e.g. space of probabilities) where all data samples can be treated seamlessly.…”

Section: Proposed Approachmentioning

confidence: 99%

Reality mining in urban space

2013

View full text Add to dashboard Cite

The widespread adoption of smartphones and other sensor-capable devices and the availability of infrastructure capable of storing and sharing sensor information have been the main enablers for the intriguing and ambitious concept of Reality Mining. Reality mining is concerned with the collection and analysis of sensed data pertaining to different aspects of human social behavior, allowing to observe the societal dynamics and predict what individuals will do next, model the behavior of large organizations, detect trends, spot emerging phenomena and events, mine public opinion, etc. The concept of reality mining is even more intriguing when applied in the context of urban space, since user-generated content can be analyzed to reveal the citizens' perspective on various aspects of the city reality. In this paper, after elaborating on the concept of reality mining in urban space, we present three cases where user-generated content is employed to discover the citizens' perspective about: a) the city attractions, b) the city issues/problems, and c) major events in the city.

show abstract

High order pLSA for indexing tagged images

Cited by 13 publications

References 52 publications

Sparse multi-modal probabilistic latent semantic analysis for single-image super-resolution

Sparse multi-modal probabilistic latent semantic analysis for single-image super-resolution

Class-specific Gaussian-multinomial latent Dirichlet allocation for image annotation

Reality mining in urban space

Contact Info

Product

Resources

About