Incremental parsing for latent semantic indexing of images

Bae, Soo Hyun; Juang, Biing‐Hwang

doi:10.1109/icip.2008.4711907

Cited by 3 publications

(2 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Also, we randomly chose 600 query images, each of which contains only one visual concept in order to minimize confusion. We follow the way of the definition of the visual concepts, the number of images, and the number of query images in each visual concept, provided in [3].…”

Section: Image Databasementioning

confidence: 99%

“…The proposed scheme parses a given multidimensional source into a number of variable-size patches, we call this methodology a parsed representation, and achieves a significantly improved performance of image compression applications. Based on the source characterization property of the proposed parsing scheme, we also proposed a query-by-example content-based image retrieval (CBIR) framework [3] and implemented image retrieval systems, called IPSILON systems. The proposed CBIR systems analyze the semantic concepts of visual information by the latent semantic analysis (LSA).…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Aspect modeling of parsed representation for image retrieval

Bae

Juang

2009

2009 IEEE International Conference on Acoustics, Speech and Signal Processing

Self Cite

View full text Add to dashboard Cite

A probabilistic framework based on a universal source coding for content-based image retrieval is proposed. By a multidimensional incremental parsing technique, which is an extension of the LempelZiv incremental parsing algorithm, a given image is parsed into a number of variable-size rectangular blocks, called parsed representations. To achieve a semantically relevant pattern matching, we introduce a new similarity measure from the first-and second-order statistics of given image patches. Once the occurrence patterns of images in the corpus are analyzed, the term-document joint distribution is estimated by an aspect modeling technique under the assumption of latent aspects. To compare the performance of the proposed image retrieval framework based on the parsed representations, we implement a benchmark system based on the fixed-shape block representations trained by vector quantization. In addition to these two systems, we bring two content-based image retrieval systems into the performance evaluation. The experimental results on a database of 20,000 natural scene images demonstrate that the proposed image retrieval system significantly outperforms other existing and the benchmark systems.

show abstract

Section: Image Databasementioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Aspect modeling of parsed representation for image retrieval

Bae

Juang

2009

2009 IEEE International Conference on Acoustics, Speech and Signal Processing

Self Cite

View full text Add to dashboard Cite

show abstract

Geospatial image mining for nuclear proliferation detection: Challenges and new opportunities

Vatsavai

Bhaduri

Cheriyadat

et al. 2010

2010 IEEE International Geoscience and Remote Sensing Symposium

View full text Add to dashboard Cite

With increasing understanding and availability of nuclear technologies, and increasing persuasion of nuclear technologies by several new countries, it is increasingly becoming important to monitor the nuclear proliferation activities. There is a great need for developing technologies to automatically or semi-automatically detect nuclear proliferation activities using remote sensing. Images acquired from earth observation satellites is an important source of information in detecting proliferation activities. High-resolution remote sensing images are highly useful in verifying the correctness, as well as completeness of any nuclear program. DOE national laboratories are interested in detecting nuclear proliferation by developing advanced geospatial image mining algorithms. In this paper we describe the current understanding of geospatial image mining techniques and enumerate key gaps and identify future research needs in the context of nuclear proliferation.Index Terms-Nuclear proliferation, low-level features, semantic classification, geospatial ontology GEOSPATIAL IMAGE MININGIncreasing resolution, volume, and availability of remote sensing imagery made it possible to accurately identify key geospatial features and their changes over time. Recent studies have shown the usefulness of remote sensing imagery for monitoring nuclear safeguards and proliferation activities [1]. Classification is one of the widely used technique for extracting thematic information. Classification is often performed *Contact: bhaduribl@ornl.gov on per-pixel basis; however proliferation detection requires identification of complex objects, patterns and their spatial relationships. One key distinguishing feature as compared to traditional thematic classification is that the objects and patterns that constitute a nuclear facility have interesting spatial relationships (metric, topological, etc) among themselves. These limitations are clearly evident from Figure 1. Classification technology is mature for extracting thematic classes such as buildings, forest, crops, etc. However, such thematic labels are not enough to capture the fact that the given image contains a nuclear power plant. What is missing is the fact that the objects (e.g., switch yard, containment building, turbine building, cooling towers) and their spatial relationships (arrangements or configurations) are not captured in traditional thematic classification. In addition, traditional image analysis approaches mainly exploit low-level image features (such as, color and texture and, to some extent, size and shape) and are oblivious to higher level descriptors and important spatial (topological) relationships without which we can not accurately discover these complex objects or higher level semantic concepts. One stumbling block in exploiting such relationships is in the description of compound objects and the spatial relationships among the object constituents. Therefore, for effective utilization of remote sensing imagery, first it is important to identify key concepts that des...

show abstract

Image analysis and compression: renewed focus on texture

Pappas¹,

Zujovic²,

Neuhoff

2010

SPIE Proceedings

View full text Add to dashboard Cite

We argue that a key to further advances in the fields of image analysis and compression is a better understanding of texture. We review a number of applications that critically depend on texture analysis, including image and video compression, content-based retrieval, visual to tactile image conversion, and multimodal interfaces. We introduce the idea of "structurally lossless" compression of visual data that allows significant differences between the original and decoded images, which may be perceptible when they are viewed side-by-side, but do not affect the overall quality of the image. We then discuss the development of objective texture similarity metrics, which allow substantial point-by-point deviations between textures that according to human judgment are essentially identical.

show abstract

Incremental parsing for latent semantic indexing of images

Cited by 3 publications

References 8 publications

Aspect modeling of parsed representation for image retrieval

Aspect modeling of parsed representation for image retrieval

Geospatial image mining for nuclear proliferation detection: Challenges and new opportunities

Image analysis and compression: renewed focus on texture

Contact Info

Product

Resources

About