Resulted word counts optimization—A new approach for better automatic image annotation

Kwaśnicka, Halina; Paradowski, Mariusz

doi:10.1016/j.patcog.2008.06.017

Cited by 8 publications

(16 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…(Paradowski 2008) and IAPR TC-12 (Grubinger et al 2006), whose characteristics are shown in Table 8. For the MGV and ICPR datasets as the reference points we have obtained the results presented in Kwasnicka and Paradowski (2008). For these data sets the proposed method achieved significantly better results.…”

Section: Patsi Results With Different Similarity Spacesmentioning

confidence: 99%

“…To obtain expressive results, 4-fold cross validation is applied and the average Fscore value is calculated. Generally the experiments were conducted using the MGV dataset (Paradowski 2008;Kwasnicka and Paradowski 2008), whose details can be found in Table 1. It is of acceptable size to allow for fast but still representative experiment execution.…”

Section: Experimental Study Of the Patsi Methodsmentioning

confidence: 99%

“…Methods of the second category employ trained classifiers to find correlations between the words and the annotated image's visual features. We can mention here Bayes Point Machine (Chang et al 2003), Support Vector Machine (Cusano et al 2004) and Decision Trees (Kwasnicka and Paradowski 2008) which all estimate the visual features distributions associated with each word. Some authors try to refine the annotation results by reducing the difference between the expected and resulting word count vectors (Kwasnicka and Paradowski 2006), by using Word-Net which contains semantic relations between words (Jin et al 2005) or by word co-occurrence models coupled with fast random walks (Llorente et al 2009), an interesting approach exploiting the recent advances in graph processing.…”

Section: Related Approachesmentioning

confidence: 99%

See 2 more Smart Citations

Image auto-annotation with automatic selection of the annotation length

2012

Self Cite

View full text Add to dashboard Cite

Developing a satisfactory and effective method for auto-annotating images that works under general conditions is a challenging task. The advantages of such a system would be manifold: it can be used to annotate existing, large databases of images, rendering them accessible to text search engines; or it can be used as core for image retrieval based on a query image's visual content. Manual annotation of images is a difficult, tedious and time consuming task. Furthermore, manual annotations tend to show great inter-person variance: considering an image, the opinions about what elements are significant and deserve an annotation vary strongly. The latter poses a problem for the evaluation of an automatic method, as an annotation's correctness is greatly subjective. In this paper we present an automatic method for annotating images, which addresses one of the existing methods' major limitation, namely a fixed annotation length. The proposed method, PATSI, automatically chooses the resulting annotation's length for each query image. It is held as simple as possible and a build-in parameter optimization procedure renders PATSI de-facto parameter free. Finally, PATSI is evaluated on standard datasets, outperforming various state-of-the-art methods.

show abstract

Section: Patsi Results With Different Similarity Spacesmentioning

confidence: 99%

Section: Experimental Study Of the Patsi Methodsmentioning

confidence: 99%

Section: Related Approachesmentioning

confidence: 99%

See 1 more Smart Citation

Image auto-annotation with automatic selection of the annotation length

2012

Self Cite

View full text Add to dashboard Cite

show abstract

“…Other model-based solutions are based e.g. on supervised topic modeling [30], supervised multi-class labeling [5], or decision trees [13]. Many other examples can be found in [34].…”

Section: Image Annotation Techniquesmentioning

confidence: 99%

Content-based annotation and classification framework

Batko

Botorek

Budikova

et al. 2013

Proceedings of the 17th International Database Engineering &Amp; Applications Symposium on - IDEAS '13

View full text Add to dashboard Cite

Unprecedented amounts of digital data are becoming available nowadays, but frequently the data lack some semantic information necessary to effectively organize these resources. For images in particular, textual annotations that represent the semantics are highly desirable. Only a small percentage of images is created with reliable annotations, therefore a lot of effort is being invested into automatic image annotation. In this paper, we address the annotation problem from a general perspective and introduce a new annotation model that is applicable to many text assignment problems. We also provide experimental results from several implemented instances of our model.

show abstract

“…Incorrect word distribution (word frequencies in the resulted annotations) in the generated annotations usually results in a large decrease of annotation quality. The problem has been addressed by many researchers [4,5,9,10,16,24,25], including the authors [19][20][21]. The authors proposed a method of modifying an automatic image annotator to match the requested word frequencies.…”

Section: Introductionmentioning

confidence: 99%

Improved Resulted Word Counts Optimizer for Automatic Image Annotation Problem

Paradowski

Kwaśnicka²

2009

Fundamenta Informaticae

Self Cite

View full text Add to dashboard Cite

Automatic Image Annotation is an important research topic in pattern recognition area. There are many different approaches to Automatic Image Annotation. In many of these approaches a key problem is determining correct word distributions in the generated annotations. Incorrect word distributions (word frequencies) results in large reduction of annotation quality. The paper presents a generic approach to find correct word frequencies. The approach may be used with many automatic image annotators, based on various machine learning paradigms. Proposed method is an improved version of our already presented method, called Greedy Resulted Word Counts Optimization. The key differences and novelties in the paper are: the concept of border support values and a method of optimal step selection in the optimization routine. Optimal step selection allows to reduce the number of computations during the optimization procedure, comparing to old Greedy Resulted Word Counts Optimization method.

show abstract

Resulted word counts optimization—A new approach for better automatic image annotation

Cited by 8 publications

References 7 publications

Image auto-annotation with automatic selection of the annotation length

Image auto-annotation with automatic selection of the annotation length

Content-based annotation and classification framework

Improved Resulted Word Counts Optimizer for Automatic Image Annotation Problem

Contact Info

Product

Resources

About