David Steele

Bio: David Steele is an academic researcher from IBM. The author has contributed to research in topics: Image segmentation & Scale-space segmentation. The author has an hindex of 5, co-authored 11 publications receiving 5611 citations.

Query by image and video content: the QBIC system

Abstract: Research on ways to extend and improve query methods for image databases is widespread. We have developed the QBIC (Query by Image Content) system to explore content-based retrieval methods. QBIC allows queries on large image and video databases based on example images, user-constructed sketches and drawings, selected color and texture patterns, camera and object motion, and other graphical information. Two key properties of QBIC are (1) its use of image and video content-computable properties of color, texture, shape and motion of images, videos and their objects-in the queries, and (2) its graphical query language, in which queries are posed by drawing, selecting and other graphical means. This article describes the QBIC system and demonstrates its query capabilities. QBIC technology is part of several IBM products. >

Query by image and video content: the QBIC system

Abstract: Research on ways to extend and improve query methods for image databases is widespread. We have developed the QBIC (Query by Image Content) system to explore content-based retrieval methods. QBIC allows queries on large image and video databases based on example images, user-constructed sketches and drawings, selected color and texture patterns, camera and object motion, and other graphical information. Two key properties of QBIC are (1) its use of image and video content-computable properties of color, texture, shape and motion of images, videos and their objects-in the queries, and (2) its graphical query language, in which queries are posed by drawing, selecting and other graphical means. This article describes the QBIC system and demonstrates its query capabilities. QBIC technology is part of several IBM products. >

A Fast Algorithm for MDL-Based Multi-Band Image Segmentation

Abstract: We consider the problem of image segmentation and describe an algorithm that is based on the Minimum Description Length (MDL) principle, is fast, is applicable to multiband images, and guarantees closed regions. We construct an objective function that, when minimized, yields a partitioning of the image into regions where the pixel values in each band of each region are described by a polynomial surface plus noise. The polynomial orders and their coefficients are determined by the algorithm. The minimization is difficult because (1) it involves a search over a very large space and (2) there is extensive computation required at each stage of the search. To address the first of these problems we use a region-merging minimization algorithm. To address the second we use an incremental polynomial regression that uses computations from the previous stage to compute results in the current stage, resulting in a significant speed up over the non-incremental technique. The segmentation result obtained is suboptimal in general but of high quality. Results on real images are shown.

Unsupervised image segmentation using the minimum description length principle

Abstract: Considers the problem of unsupervised multiband image segmentation specifically, using the MDL criterion, an associated complexity measure. According to the MDL principle, the best estimates are those that result in the most compact encoding of the image. An important advantage of such an approach is that it is virtually free of the need to choose arbitrary thresholds, which are typical of many segmentation techniques, and which, in many cases, need to be interactively adjusted in order to get satisfactory results. >

Recent applications of IBM's query by image content (QBIC)

Abstract: The QBIC project at IBM's Almaden Research Center is studying methods to query large on-line image databases using the images' content as the basis of the queries. Examples of the content include color, color layout, texture, shape, size, orientation, and position of image objects and regions. We have developed an AIX prototype, WWW QBIC server with WWW demo, and a product called Ultimedia Manager (OS/2 and Windows). The key applications of this technology are in the areas where image patterns are the basis of the queries, and where keywords are not precise enough for satisfactory retrieval, such as in on-line stock photo, retail, art etc. In this paper we briefly outline several recent applications of QBIC. SEDONA Systems Corporation, PIP Printing and IBM announced an application where QBIC is used to help the user find images to be included in documents remotely printed in high resolution and quality; IBM in the UK is developing an "advisor" to textile designers; Prof. B. Holt (UC Davis) researches use of QBIC for retrieving art images based on examples of a motifs used across geography and time without having to use standard authority lists of text reference terms, especially when searching for visual elements that are difficult to describe in words; and Prof. J. Hethorn (UC Davis) uses Ultimedia Manager to study colors, styles and trends in fashion. In this paper we "Permission to make digital/hard copy of all or part of this material without fee is granted provided that copies are not made or distributed for profit or commercial advantage, the ACM copyright/server notice, the title of the publication and its date appear, and notice is given that copying is by permission of the Association for Computing Machinery, Ine.(ACM). To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee." present a short overview of QBIC technology and products, and summarize the above applications. Introduction to Query by Image Content (QBIC) Today's technology enables users to acquire, store, manipulate, and transmit large numbers of images. However, finding the right image in a large image database is a difficult problem. Traditional database retrieval technology uses an "exact match" approach, which works well for traditional structured information. However, for multimedia data types, such as digitized images, this approach breaks down. More and more we are witnessing applications where very large image collections are available and where exact search is becoming increasingly inadequate. Some examples of these applications are in retail cataloging, stock photography and large collections of art and graphics. The current approach is to index and then search the images based on some identifying text such as title, descriptive keywords (from a limited vocabulary), or catalog numbers. Some programs offer full relational database (SQL) search over relational tables, where each column can contain text or numeric data. These systems are useful in many applications where exact keywords and numeric descriptors can be easily obtained and are sufficient to find images the user wants. However, more powerful searches will be necessary. We have developed a way to index and search images by the content of the image. Database queries can select for color, texture, shape and position (1,2,3,4). This approach is called QBIC, for Query by Image Content. Ultimedia Manager is an OS/2 and Windows product from IBM that combines QBIC technology with traditional database searching, and offers a powerful image browser. This unique combination provides new ways to manage image databases. The QBIC portion of Ultimedia Manager can index and retrieve images according to the following kinds of information: Average color; Histogram color (based on 64 predefined colors); Texture or Pattern; Shape (of user outline); and Position. These measures can be calculated for the whole image or for user-defined areas within the image. These areas are considered to be "visual indexes" to the image. Part of the classification (indexing) process is for the user to identify areas of interest within the image. This, for example, allows a user to search for a red flower in an image of green foliage. When running a query, Ultimedia Manager ranks each image based on a distance measurement computed between the query attribute values and the corresponding features or attributes stored in the QBIC Catalog database. This is done only for images that were selected by exact data search (that is, SQL). Users can modify the relative weights of each visual attribute to satisfy their preferences in influencing the final sorting of result images. See Fig. 1 that shows SQL data selector, color histogram selector, query window and result area window of Ultimedia Manager, in an application of searching through fashion catalogs. QBIC has an increasing importance in the new world of WWW. There is a strong interest among content providers and retailers to provide their collections (for sale or for browsing) over the WWW. Now, imagine having a large on-line collection of photographs and posing a keyword/parametric query such as "give me all SHIRTS made of COTTON in the PRICE range 10 to 20 dollars". This query may produce large number of responses, which makes it impractical for browsing since thumbnails are obtained through standard WWW connections (modems). (After all, we do not buy shirts by only specifying price and material either). Thus, one of the basic functions of image searching, namely fast and easy to use browsing, can not be performed on the WWW due to low ly~ndwidth. Now, imagine that in applications where image appearance is important (on-line stock photo, retail, fashion, art) one can specify some kind of QBIC query, like a set of colors. QBIC, running at the server, can sort the images that satisfy the initial keyword/parametric query and thert send to the browser only a few top matches (in this case having the desired set of colors), thus dramatically reducing the bandwidth requirements. A user can either be satisfied with this or use relevance feedback to change colors or by clicking at the image to ask for "more images like this". This way an easy to use and fast search/browse is possible even in the case of slow WWW access. We have developed a QBIC server for WWW with associated APIs and simple GUI, that has been operational since mid 1995 in the form of a demo with 1900 images. It offers color histogram query (selection of desired colors and their amounts) and a new very powerful color layout query which is ideally suited for on-line stock photography (setection of colors and their position in the image by using simple painting program). The current interface uses standard HTML and plans are under way to make it more interactive. This server, part of several IBM customer solutions, is part of the IBM Digital Library where it is integrated with text and parametric search, and is also available as a special deal. Fig. 2 shows our WWW QBIC browser. QBIC applications in stock-photography for remote printing As a result of an agreement between SEDONA Systems Corporation, PIP Printing, and IBM. desktop publishers, graphic designers, and consumers can now quickly and inexpensively search for and print thousands of high-quality, royalty-free images via the lnternet, using QBIC technology. The agreement will enable Internet users to search for stock images and clip art by selecting attributes including color, texture, and shape using IBM°s award-winning Query by Image Content (QBIC) technology. These images can be quickly transmitted to loeat or distant PIP Printing centers using REPRINT, SEDONA's document communications software. The companies announced the new service at the Seybold '95 Conference. Stock images and clip art may be retrieved from the Internet without royalty payments, then resized, rotated, and embedded in

Data Mining: Concepts and Techniques

Abstract: The increasing volume of data in modern business and science calls for more complex and sophisticated tools. Although advances in data mining technology have made extensive data collection much easier, it's still always evolving and there is a constant need for new techniques and tools that can help us transform this data into useful information and knowledge. Since the previous edition's publication, great advances have been made in the field of data mining. Not only does the third of edition of Data Mining: Concepts and Techniques continue the tradition of equipping you with an understanding and application of the theory and practice of discovering patterns hidden in large data sets, it also focuses on new, important topics in the field: data warehouses and data cube technology, mining stream, mining social networks, and mining spatial, multimedia and other complex data. Each chapter is a stand-alone guide to a critical topic, presenting proven algorithms and sound implementations ready to be used directly or with strategic modification against live data. This is the resource you need if you want to apply today's most powerful data mining techniques to meet real business challenges. * Presents dozens of algorithms and implementation examples, all in pseudo-code and suitable for use in real-world, large-scale data mining projects. * Addresses advanced topics such as mining object-relational databases, spatial databases, multimedia databases, time-series databases, text databases, the World Wide Web, and applications in several fields. *Provides a comprehensive, practical look at the concepts and techniques you need to get the most out of real business data

Content-based image retrieval at the end of the early years

Abstract: Presents a review of 200 references in content-based image retrieval. The paper starts with discussing the working conditions of content-based retrieval: patterns of use, types of pictures, the role of semantics, and the sensory gap. Subsequent sections discuss computational steps for image retrieval systems. Step one of the review is image processing for retrieval sorted by color, texture, and local geometry. Features for retrieval are discussed next, sorted by: accumulative and global features, salient points, object and shape features, signs, and structural combinations thereof. Similarity of pictures and objects in pictures is reviewed for each of the feature types, in close connection to the types and means of feedback the user of the systems is capable of giving by interaction. We briefly discuss aspects of system engineering: databases, system architecture, and evaluation. In the concluding section, we present our view on: the driving force of the field, the heritage from computer vision, the influence on computer vision, the role of similarity and of interaction, the need for databases, the problem of evaluation, and the role of the semantic gap.

Approximate nearest neighbors: towards removing the curse of dimensionality

Abstract: We present two algorithms for the approximate nearest neighbor problem in high-dimensional spaces. For data sets of size n living in R d , the algorithms require space that is only polynomial in n and d, while achieving query times that are sub-linear in n and polynomial in d. We also show applications to other high-dimensional geometric problems, such as the approximate minimum spanning tree. The article is based on the material from the authors' STOC'98 and FOCS'01 papers. It unifies, generalizes and simplifies the results from those papers.

Computer Vision: Algorithms and Applications

Abstract: Humans perceive the three-dimensional structure of the world with apparent ease. However, despite all of the recent advances in computer vision research, the dream of having a computer interpret an image at the same level as a two-year old remains elusive. Why is computer vision such a challenging problem and what is the current state of the art? Computer Vision: Algorithms and Applications explores the variety of techniques commonly used to analyze and interpret images. It also describes challenging real-world applications where vision is being successfully used, both for specialized applications such as medical imaging, and for fun, consumer-level tasks such as image editing and stitching, which students can apply to their own personal photos and videos. More than just a source of recipes, this exceptionally authoritative and comprehensive textbook/reference also takes a scientific approach to basic vision problems, formulating physical models of the imaging process before inverting them to produce descriptions of a scene. These problems are also analyzed using statistical models and solved using rigorous engineering techniques Topics and features: structured to support active curricula and project-oriented courses, with tips in the Introduction for using the book in a variety of customized courses; presents exercises at the end of each chapter with a heavy emphasis on testing algorithms and containing numerous suggestions for small mid-term projects; provides additional material and more detailed mathematical topics in the Appendices, which cover linear algebra, numerical techniques, and Bayesian estimation theory; suggests additional reading at the end of each chapter, including the latest research in each sub-field, in addition to a full Bibliography at the end of the book; supplies supplementary course material for students at the associated website, http://szeliski.org/Book/. Suitable for an upper-level undergraduate or graduate-level course in computer science or engineering, this textbook focuses on basic techniques that work under real-world conditions and encourages students to push their creative boundaries. Its design and exposition also make it eminently suitable as a unique reference to the fundamental techniques and current research literature in computer vision.

Similarity Search in High Dimensions via Hashing

Abstract: The nearestor near-neighbor query problems arise in a large variety of database applications, usually in the context of similarity searching. Of late, there has been increasing interest in building search/index structures for performing similarity search over high-dimensional data, e.g., image databases, document collections, time-series databases, and genome databases. Unfortunately, all known techniques for solving this problem fall prey to the \curse of dimensionality." That is, the data structures scale poorly with data dimensionality; in fact, if the number of dimensions exceeds 10 to 20, searching in k-d trees and related structures involves the inspection of a large fraction of the database, thereby doing no better than brute-force linear search. It has been suggested that since the selection of features and the choice of a distance metric in typical applications is rather heuristic, determining an approximate nearest neighbor should su ce for most practical purposes. In this paper, we examine a novel scheme for approximate similarity search based on hashing. The basic idea is to hash the points Supported by NAVY N00014-96-1-1221 grant and NSF Grant IIS-9811904. Supported by Stanford Graduate Fellowship and NSF NYI Award CCR-9357849. Supported by ARO MURI Grant DAAH04-96-1-0007, NSF Grant IIS-9811904, and NSF Young Investigator Award CCR9357849, with matching funds from IBM, Mitsubishi, Schlumberger Foundation, Shell Foundation, and Xerox Corporation. Permission to copy without fee all or part of this material is granted provided that the copies are not made or distributed for direct commercial advantage, the VLDB copyright notice and the title of the publication and its date appear, and notice is given that copying is by permission of the Very Large Data Base Endowment. To copy otherwise, or to republish, requires a fee and/or special permission from the Endowment. Proceedings of the 25th VLDB Conference, Edinburgh, Scotland, 1999. from the database so as to ensure that the probability of collision is much higher for objects that are close to each other than for those that are far apart. We provide experimental evidence that our method gives signi cant improvement in running time over other methods for searching in highdimensional spaces based on hierarchical tree decomposition. Experimental results also indicate that our scheme scales well even for a relatively large number of dimensions (more than 50).