Interactive browsing via diversified visual summarization for image search results

Presenting and browsing image search results play key roles in helping users to find desired images from search results. Most existing commercial image search engines present them, dependent on a ranked list. However, such a scheme suffers from at least two drawbacks: inconvenience for consumers to get an overview of the whole result, and high computation cost to find desired images from the list. In this paper, we introduce a novel search result summarization approach and exploit this approach to further propose an interactive browsing scheme. The main contribution of this paper includes: (1) a dynamic absorbing random walk to find diversified representatives for image search result summarization; (2) a local scaled visual similarity evaluation scheme between two images through inspecting the relation between each image and other images; and (3) an interactive browsing scheme, based on a tree structure for organizing the images obtained from the summarization approach, to enable users to intuitively and conveniently browse the image search results. Quantitative experimental results and user study demonstrate the effectiveness of the proposed summarization and browsing approaches.     

Multiple feature kernel hashing for large-scale visual search

Recently hashing has become attractive in large-scale visual search, owing to its theoretical guarantee and practical success. However, most of the state-of-the-art hashing methods can only employ a single feature type to learn hashing functions. Related research on image search, clustering, and other domains has proved the advantages of fusing multiple features. In this paper we propose a novel multiple feature kernel hashing framework, where hashing functions are learned to preserve certain similarities with linearly combined multiple kernels corresponding to different features. The framework is not only compatible with general types of data and diverse types of similarities indicated by different visual features, but also general for both supervised and unsupervised scenarios. We present efficient alternating optimization algorithms to learn both the hashing functions and the optimal kernel combination. Experimental results on three large-scale benchmarks CIFAR-10, NUS-WIDE and a-TRECVID show that the proposed approach can achieve superior accuracy and efficiency over state-of-the-art methods.     

A next generation browsing environment for large image repositories

Next generation environments will change the way people work and live as they will provide new advances in areas ranging from remote work and education, e-commerce, gaming to information-on-demand. In many of these applications intelligent interpretation of multimedia data such as image, video and audio resources is necessary. In this paper we present an effective approach to handling image repositories providing the user with an intuitive interface of visualising and browsing large collections of pictures. Based on the idea of similarity-based organisation of images where images that are visually similar are located close to each other in visualisation space, images are projected onto a sphere with which the user can interact. Rotating the sphere reveals images of different colours while tilting operations focus on brighter or darker images. Large image collections are handled through a hierarchical approach that brings up similar, previously hidden, images when zooming in on an area. Furthermore, the way images are organised can be interactively changed by the user. Our next generation browsing environment has been successfully tested on a large database of several thousand images.     

Using visual spatial search interface for WWW applications

Most Internet search engines are keyword-based. They are not efficient for the queries where geographical location is important, such as finding hotels within an area or close to a place of interest. A natural interface for spatial searching is a map, which can be used not only to display locations of search results but also to assist forming search conditions. A map-based search engine requires a well-designed visual interface that is intuitive to use yet flexible and expressive enough to support various types of spatial queries as well as aspatial queries. Similar to hyperlinks for text and images in an HTML page, spatial objects in a map should support hyperlinks. Such an interface needs to be scalable with the size of the geographical regions and the number of websites it covers. In spite of handling typically a very large amount of spatial data, a map-based search interface should meet the expectation of fast response time for interactive applications. In this paper we discuss general requirements and the design for a new map-based web search interface, focusing on integration with the WWW and visual spatial query interface. A number of current and future research issues are discussed, and a prototype for the University of Queensland is presented.     

Genre identification for office document search and browsing

When searching or browsing documents, the genre of a document is an important consideration that complements topical characterization. We examine design considerations for automatic tagging of office document pages with genre membership. These include selecting features that characterize genre-related information in office documents, examining the utility of text-based features and image-based features, and proposing a simple ensemble method to improve the performance of genre identification. Experiments were conducted on the open-set identification of four coarse office document genres: technical paper, photo, slide, and table. Our experiments show that when combined with image-based features, text-based features do not significantly influence performance. These results provide support for a topic-independent approach to identification of coarse office document genres. Experiments also show that our simple ensemble method significantly improves performance relative to using a support vector machine (SVM) classifier alone. We demonstrate the utility of our approach by integrating our automatic genre tags in a faceted search and browsing application for office document collections.     

Interactive unsupervised classification and visualization for browsing an image collection

In this paper, we propose an approach to interactive navigation in image collections. As structured groups are more appealing to users than flat image collections, we propose an image clustering algorithm, with an incremental version that handles time-varying collections. A 3D graph-based visualization technique reflects the classification state. While this classification visualization is itself interactive, we show how user feedback may assist the classification, thus enabling a user to improve it.     

Effective and efficient similarity search in scientific workflow repositories

Scientific workflows have become a valuable tool for large-scale data processing and analysis. This has led to the creation of specialized online repositories to facilitate workflow sharing and reuse. Over time, these repositories have grown to sizes that call for advanced methods to support workflow discovery, in particular for similarity search. Effective similarity search requires both high quality algorithms for the comparison of scientific workflows and efficient strategies for indexing, searching, and ranking of search results. Yet, the graph structure of scientific workflows poses severe challenges to each of these steps. Here, we present a complete system for effective and efficient similarity search in scientific workflow repositories, based on the Layer Decomposition approach to scientific workflow comparison. Layer Decomposition specifically accounts for the directed dataflow underlying scientific workflows and, compared to other state-of-the-art methods, delivers best results for similarity search at comparably low runtimes. Stacking Layer Decomposition with even faster, structure-agnostic approaches allows us to use proven, off-the-shelf tools for workflow indexing to further reduce runtimes and scale similarity search to sizes of current repositories.     

Abstract picture generation and zooming user interface for intuitive music browsing

Today many people store music media files in personal computers or portable audio players, thanks to recent evolution of multimedia technologies. The more music media files these devices store, the messier it is to search for tunes that users want to listen to. We propose MusCat, a music browser to interactively search for the tunes according to features, not according to metadata (e.g. title, artist name). The technique firstly calculates features of tunes, and then hierarchically clusters the tunes according to the features. It then automatically generates abstract pictures, so that users can recognize characteristics of tunes more instantly and intuitionally. It finally visualizes the tunes by using abstract pictures. MusCat enables intuitive music selection with the zooming user interface.     

Interactive visual modeling for performance

An overview is given of Q+, an interactive tool for performance modeling that uses graphical input and visual output. Two major enhancements are a subnetwork capability for structuring models hierarchically and an integrated expression capability. New capabilities are custom icons and temporal browsing. With a Q+ icon palette, users can draw their own icons and manipulate existing ones. The browser allows browsing, editing and updating Q+ information, which can be textual or graphical. Automatic model building, operations management, and experimental design with Q+ are discussed     

ViBE: a compressed video database structured for active browsing and search

In this paper, we describe a unique new paradigm for video database management known as ViBE (video indexing and browsing environment). ViBE is a browseable/searchable paradigm for organizing video data containing a large number of sequences. The system first segments video sequences into shots by using a new feature vector known as the Generalized Trace obtained from the DC-sequence of the compressed data. Each video shot is then represented by a hierarchical structure known as the shot tree. The shots are then classified into pseudo-semantic classes that describe the shot content. Finally, the results are presented to the user in an active browsing environment using a similarity pyramid data structure. The similarity pyramid allows the user to view the video database at various levels of detail. The user can also define semantic classes and reorganize the browsing environment based on relevance feedback. We describe how ViBE performs on a database of MPEG sequences.     

Miniatures versus icons as a visual cache for videotex browsing

Miniatures are an alternative to icons for the representation of a large graphical object such as a window in a reduced format. A front end user interface to an existing videotex system was implemented using icons as well as miniatures to represent previously seen frames in a visual cache, and an empirical comparison showed that users had the same performance with the two representations but subjectively preferred icons.     

Special issue on towards advancements in machine learning for exploiting large-scale and heterogeneous repositories

Neural Computing and Applications -     

Isoresponse time regions for the evaluation of visual search performance in ergonomic interface models.

Computer-aided ergonomic evaluation models often provide a limited function of visual performance, usually in the form of visibility tests at any given location. It is, however, of great importance to complete a visual task in a predetermined time successfully or to predict the time to complete the task, e.g., a task of high-level control such as aircraft flight operations. In this study, isoresponse time regions which have an effect on visual performance in real-life situations are presented as an effort to complement aforementioned shortcomings of ergonomic interface models. A multiple regression was used to provide equal response time regions, being asymmetrical in shape. In addition, the fuzzy linear regression (FLR) model was introduced to provide a fuzzy set from which a decision-maker might select appropriately. It provided similar fitted values with having the advantage of utilizing the interval of those fitted values over the conventional regression model. The results can be applied effectively to the design of an interface for diverse visual tasks in real situations.     

Perceptual training for visual search

People are better at visual search than the best fully automated methods. Despite this, visual search remains a difficult perceptual task. The goal of this investigation was to experimentally test the ways in which visual search performance could be improved through two categories of training interventions: perceptual training and conceptual training. To determine the effects of each training on a later performance task, the two types of trainings were manipulated using a between-subjects design (conceptual vs. perceptual × training present vs. training absent). Perceptual training led to speed and accuracy improvements in visual search. Issues with the design and administration of the conceptual training limited conclusions on its effectiveness but provided useful lessons for conceptual training design. The results suggest that when the visual search task involves detecting heterogeneous or otherwise unpredictable stimuli, perceptual training can improve visual search performance. Similarly, careful consideration of the performance task and training design is required to evaluate the effectiveness of conceptual training.

Practitioner Summary: Visual search is a difficult, yet critical, task in industries such as baggage screening and radiology. This study investigated the effectiveness of perceptual training for visual search. The results suggest that when visual search involves detecting heterogeneous or otherwise unpredictable stimuli, perceptual training may improve the speed and accuracy of visual search.     

Very large-scale neighborhood search

Neighborhood search algorithms are often the most effective approaches available for solving partitioning problems, a difficult class of combinatorial optimization problems arising in many application domains including vehicle routing, telecommunications network design, parallel machine scheduling, location theory, and clustering. A critical issue in the design of a neighborhood search algorithm is the choice of the neighborhood structure, that is, the manner in which the neighborhood is defined. Currently, the two-exchange neighborhood is the most widely used neighborhood for solving partitioning problems. The paper describes the cyclic exchange neighborhood , which is a generalization of the two-exchange neighborhood in which a neighbor is obtained by performing a cyclic exchange . The cyclic exchange neighborhood has substantially more neighbors compared to the two-exchange neighborhood. This paper outlines a network optimization based methodology to search the neighborhood efficiently and presents a proof of concept by applying it to the capacitated minimum spanning tree problem, an important problem in telecommunications network design.     

A visual backchannel for large-scale events

We introduce the concept of a Visual Backchannel as a novel way of following and exploring online conversations about large-scale events. Microblogging communities, such as Twitter, are increasingly used as digital backchannels for timely exchange of brief comments and impressions during political speeches, sport competitions, natural disasters, and other large events. Currently, shared updates are typically displayed in the form of a simple list, making it difficult to get an overview of the fast-paced discussions as it happens in the moment and how it evolves over time. In contrast, our Visual Backchannel design provides an evolving, interactive, and multi-faceted visual overview of large-scale ongoing conversations on Twitter. To visualize a continuously updating information stream, we include visual saliency for what is happening now and what has just happened, set in the context of the evolving conversation. As part of a fully web-based coordinated-view system we introduce Topic Streams, a temporally adjustable stacked graph visualizing topics over time, a People Spiral representing participants and their activity, and an Image Cloud encoding the popularity of event photos by size. Together with a post listing, these mutually linked views support cross-filtering along topics, participants, and time ranges. We discuss our design considerations, in particular with respect to evolving visualizations of dynamically changing data. Initial feedback indicates significant interest and suggests several unanticipated uses.