Semantic Based Visual Information Retrieval

"This book presents state-of-the-art advancements and developments in the field, and also brings a selection of techniques and algorithms about semantic-based visual information retrieval. It covers many critical issues, such as: multi-level representation and description, scene understanding, semantic modeling, image and video annotation, human-computer interaction, and more"--Provided by publisher.

Visual information retrieval (VIR) is an active and vibrant research area, which attempts at providing means for organizing, indexing, annotating, and retrieving visual information (images and videos) from large, unstructured repositories. The goal of VIR is to retrieve matches ranked by their relevance to a given query, which is often expressed as an example image and/or a series of keywords. During its early years (1995-2000), the research efforts were dominated by content-based approaches contributed primarily by the image and video processing community. During the past decade, it was widely recognized that the challenges imposed by the lack of coincidence between an image's visual contents and its semantic interpretation, also known as semantic gap, required a clever use of textual metadata (in addition to information extracted from the image's pixel contents) to make image and video retrieval solutions efficient and effective. The need to bridge (or at least narrow) the semantic gap has been one of the driving forces behind current VIR research. Additionally, other related research problems and market opportunities have started to emerge, offering a broad range of exciting problems for computer scientists and engineers to work on. In this introductory book, we focus on a subset of VIR problems where the media consists of images, and the indexing and retrieval methods are based on the pixel contents of those images -- an approach known as content-based image retrieval (CBIR). We present an implementation-oriented overview of CBIR concepts, techniques, algorithms, and figures of merit. Most chapters are supported by examples written in Java, using Lucene (an open-source Java-based indexing and search implementation) and LIRE (Lucene Image REtrieval), an open-source Java-based library for CBIR. Table of Contents: Introduction / Information Retrieval: Selected Concepts and Techniques / Visual Features / Indexing Visual Features / LIRE: An Extensible Java CBIR Library / Concluding Remarks

Content-based Image Retrieval (CBIR) ist ein Verfahren zum Auffinden von Bildern in großen Datenbanken wie z. B. dem Internet anhand ihres Inhalts. Ausgehend von einem vom Nutzer bereitgestellten Anfragebild, gibt das System eine sortierte Liste ähnlicher Bilder zurück. Der Großteil moderner CBIR-Systeme vergleicht Bilder ausschließlich anhand ihrer visuellen Ähnlichkeit, d.h. dem Vorhandensein ähnlicher Texturen, Farbkompositionen etc. Jedoch impliziert visuelle Ähnlichkeit nicht zwangsläufig auch semantische Ähnlichkeit. Zum Beispiel können Bilder von Schmetterlingen und Raupen als ähnlich betrachtet werden, weil sich die Raupe irgendwann in einen Schmetterling verwandelt. Optisch haben sie jedoch nicht viel gemeinsam. Die vorliegende Arbeit stellt eine Methode vor, welche solch menschliches Vorwissen über die Semantik der Welt in Deep-Learning-Verfahren integriert. Als Quelle für dieses Wissen dienen Taxonomien, die für eine Vielzahl von Domänen verfügbar sind und hierarchische Beziehungen zwischen Konzepten kodieren (z.B., ein Pudel ist ein Hund ist ein Tier etc.). Diese hierarchiebasierten semantischen Bildmerkmale verbessern die semantische Konsistenz der CBIR-Ergebnisse im Vergleich zu herkömmlichen Repräsentationen und Merkmalen erheblich. Darüber hinaus werden drei verschiedene Mechanismen für interaktives Image Retrieval präsentiert, welche die den Anfragebildern inhärente semantische Ambiguität durch Einbezug von Benutzerfeedback auflösen. Eine der vorgeschlagenen Methoden reduziert das erforderliche Feedback mithilfe von Clustering auf einen einzigen Klick, während eine andere den Nutzer kontinuierlich involviert, indem das System aktiv nach Feedback zu denjenigen Bildern fragt, von denen der größte Erkenntnisgewinn bezüglich des Relevanzmodells erwartet wird. Die dritte Methode ermöglicht dem Benutzer die Auswahl besonders interessanter Bildbereiche zur Fokussierung der Ergebnisse. Diese Techniken liefern bereits nach wenigen Feedbackrunden deutlich relevantere Ergebnisse, was die Gesamtmenge der abgerufenen Bilder reduziert, die der Benutzer überprüfen muss, um relevante Bilder zu finden. Content-based image retrieval (CBIR) aims for finding images in large databases such as the internet based on their content. Given an exemplary query image provided by the user, the retrieval system provides a ranked list of similar images. Most contemporary CBIR systems compare images solely by means of their visual similarity, i.e., the occurrence of similar textures and the composition of colors. However, visual similarity does not necessarily coincide with semantic similarity. For example, images of butterflies and caterpillars can be considered as similar, because the caterpillar turns into a butterfly at some point in time. Visually, however, they do not have much in common. In this work, we propose to integrate such human prior knowledge about the semantics of the world into deep learning techniques. Class hierarchies serve as a source for this knowledge, which are readily available for a plethora of domains and encode is-a relationships (e.g., a poodle is a dog is an animal etc.). Our hierarchy-based semantic embeddings improve the semantic consistency of CBIR results substantially compared to conventional image representations and features. We furthermore present three different mechanisms for interactive image retrieval by incorporating user feedback to resolve the inherent semantic ambiguity present in the query image. One of the proposed methods reduces the required user feedback to a single click using clustering, while another keeps the human in the loop by actively asking for feedback regarding those images which are expected to improve the relevance model the most. The third method allows the user to select particularly interesting regions in images. These techniques yield more relevant results after a few rounds of feedback, which reduces the total amount of retrieved images the user needs to inspect to find relevant ones.

"This 10-volume compilation of authoritative, research-based articles contributed by thousands of researchers and experts from all over the world emphasized modern issues and the presentation of potential opportunities, prospective solutions, and future directions in the field of information science and technology"--Provided by publisher.

"This set of books represents a detailed compendium of authoritative, research-based entries that define the contemporary state of knowledge on technology"--Provided by publisher.

Modern society exists in a digital era in which high volumes of multimedia information exists. To optimize the management of this data, new methods are emerging for more efficient information retrieval. Web Semantics for Textual and Visual Information Retrieval is a pivotal reference source for the latest academic research on embedding and associating semantics with multimedia information to improve data retrieval techniques. Highlighting a range of pertinent topics such as automation, knowledge discovery, and social networking, this book is ideally designed for researchers, practitioners, students, and professionals interested in emerging trends in information retrieval.

This book constitutes the thoroughly refereed proceedings of the 10th International Conference on Visual Information Systems, VISUAL 2008, held in Salerno, Italy, September 11-12, 2008. The 35 papers presented in this volume, together with 3 keynote speeches, were carefully reviewed and selected from 58 submissions. The topics covered are information and data visualization; advances techniques for visual information management; mobile visual information systems; image and video indexing and retrieval; applications of visual information systems; and industrial experiences.

The book describes several techniques used to bridge the semantic gap and reflects on recent advancements in content-based image retrieval (CBIR). It presents insights into and the theoretical foundation of various essential concepts related to image searches, together with examples of natural and texture image types. The book discusses key challenges and research topics in the context of image retrieval, and provides descriptions of various image databases used in research studies. The area of image retrieval, and especially content-based image retrieval (CBIR), is a very exciting one, both for research and for commercial applications. The book explains the low-level features that can be extracted from an image (such as color, texture, shape) and several techniques used to successfully bridge the semantic gap in image retrieval, making it a valuable resource for students and researchers interested in the area of CBIR alike.