Effective Semantic Annotation by Image-to-Concept Distribution Model

Su, Ja-Hwung; Chou, Chien-Li; Lin, Ching‐Yung; Tseng, Vincent S.

doi:10.1109/tmm.2011.2129502

Cited by 37 publications

(15 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In many applications on image retrieval, in advance, all training pictures were manually labeled what there are in a picture [1,2]. And these symbols are regarded as the semantic of picture.…”

Section: How To Fulfill a More Detailed Definition Of An Object?mentioning

confidence: 99%

A Bio-Inspired Integration Method for Object Semantic Representation

Wei

2016

Journal of Artificial Intelligence and Soft Computing Research

View full text Add to dashboard Cite

We have two motivations. Firstly, semantic gap is a tough problem puzzling almost all sub-fields of Artificial Intelligence. We think semantic gap is the conflict between the abstractness of high-level symbolic definition and the details, diversities of low-level stimulus. Secondly, in object recognition, a pre-defined prototype of object is crucial and indispensable for bi-directional perception processing. On the one hand this prototype was learned from perceptional experience, and on the other hand it should be able to guide future downward processing. Human can do this very well, so physiological mechanism is simulated here. We utilize a mechanism of classical and non-classical receptive field (nCRF) to design a hierarchical model and form a multi-layer prototype of an object. This also is a realistic definition of concept, and a representation of denoting semantic. We regard this model as the most fundamental infrastructure that can ground semantics. Here a AND-OR tree is constructed to record prototypes of a concept, in which either raw data at low-level or symbol at high-level is feasible, and explicit production rules are also available. For the sake of pixel processing, knowledge should be represented in a data form; for the sake of scene reasoning, knowledge should be represented in a symbolic form. The physiological mechanism happens to be the bridge that can join them together seamlessly. This provides a possibility for finding a solution to semantic gap problem, and prevents discontinuity in low-order structures.

show abstract

Section: How To Fulfill a More Detailed Definition Of An Object?mentioning

confidence: 99%

A Bio-Inspired Integration Method for Object Semantic Representation

Wei

2016

Journal of Artificial Intelligence and Soft Computing Research

View full text Add to dashboard Cite

show abstract

“…The result of this comparison yields a probability value of each keyword being present in an image. The block diagram of typical image annotation framework [7] is shown in following figure.…”

Section: Automated Image Annotationmentioning

confidence: 99%

Multi-Relation Image Retrieval and Annotation based on Holistic Approach & Decision Tree

More¹,

Mishra²

2012

IJCA

View full text Add to dashboard Cite

This work of multi-relation image retrieval and annotation is based on the holistic approach and the decision tree. In this we have proposed that for the retrieval of similar images as that of query image Dominant color descriptor (DCD) is used, this descriptor uses the color feature for the retrieval of images. This creates the feature vector index. Test keywords are correlated with the feature vector index, the correlation is performed by multi class association to get the classes for processing on them. Classes which are not necessary are discarded using cross validation in the decision tree process Decision tree used to take the relevant classes and finally we calculate the Gain of feature vector and we get the retrieval of the images based on the query image with the associated keywords or annotation. This work has been implemented on MatLab 7.5 simulator.

show abstract

“…Many multi-label learning algorithms, such as IBLR_ML [3], RAkEL [4] and Rank-CVM [5], have been witnessed during the past decades, and widely applied in many domains, including text categorization [6], image and video annotation [7], content annotation [8], music processing [9], bioinformatics [10], and so on. Generally they can be grouped into two categories, i.e., algorithm adaption and problem transformation [1,11].…”

Section: Introductionmentioning

confidence: 99%