Catálogo de publicaciones - libros

In this demo, we present the use of the ARIA platform for modular design of media processing and retrieval applications. ARIA is a middleware for describing and executing media processing workflows to process, filter, and fuse sensory inputs and actuate responses in real-time. ARIA is designed with the goal of maximum modularity and ease of integration of a diverse collection of media processing components and data sources. Moreover, ARIA is cognizant of the fact that various media operators and data structures are adaptable in nature; i.e, the delay, size, and quality/precision characteristics of these operators can be controlled via various parameters. In this demo, we present the ARIA design interface in different image processing and retrieval scenarios.

There exist various methods for stereoscopic viewing of images, most requiring that a viewer wears some special glasses. Recent technology developments have resulted in displays that enable 3D viewing without glasses. In this paper, we present results from our proposed approach to automatic rectification of two images of the same scene captured by cameras at general positions, so that the results can be viewed on a 3D display. Both simulated and real data experiments are presented.

In this paper, we provide a brief overview of the human movement analysis research at the Arts, Media and Engineering program, Arizona State University, and its applications in interactive dance. A family of robust algorithms has been developed to analyze dancers’ movement at multiple temporal and spatial levels from a number of perspectives such as marker distributions, joint angles, body silhouettes as well as weight distributions to conduct reliable dancer tracking, posture and gesture recognition. Multiple movement sensing modalities have been used and sometimes fused in our current research, including marker-based motion capture system, pressure sensitive floor and video cameras. Some of the developed algorithms have been successfully used in real life dance performances.

We present a system for visualizing event detection in video and revealing the algorithmic and scientific insights. Visual events are viewed as evolving temporal patterns in the semantic concept space. For video clips of different events, we present their corresponding traces in the semantic concept space as the event evolves. The presentation of the event in the concept space is scored by pre-trained models of the dynamics of each concept in the context of the event, which provides a measure of how well the given event matches the evolution pattern of the target event in the multi-dimensional concept space. Scores obtained for different videos is shown to project them into different parts of the final score space. This presentation walks the user through the entire process of concept-centered event recognition for events such as exiting car , riot , and airplane flying .

The VideoSOM sytem is a tool for content-based video navigation based on a growing self-organizing map. Our interface allows the user to browse the video content using simultaneously several perspectives, temporal as well as content-based representations of the video. Combined with the interaction possibilities between them this allows for efficient searching of relevant information in video content.

The growth of digital image collections in many areas of science and commercial environments has over the last decade spawned great interest in content-based image retrieval. A great variety of methods have been developed to retrieve images based on example queries and techniques to elicit and utilize relevance feedback (e.g. [4, 5]). Often the systems provide a simple framework that permits evaluation of the method and are not intended as full-fledged systems ready to be used in a realistic setting. Also, comparatively little effort has been expended on devising efficient techniques to browse image collections effectively. The few notable exceptions [1, 3, 6] treat browsing internally as a sequence of queries and thus do not leverage the performance benefits associated with pre-computed browsing structures.

We introduce an interface for efficient video search that exploits the human ability to quickly scan visual content, after automatic retrieval has arrange the images in expected order of relevance. While extreme video retrieval is taxing to the human, it is also extremely effective. Two variants of extreme retrieval are demonstrated, 1) RSVP which automatically pages through images with user-control of the paging speed, while the user marks relevant shots and 2) MBRP where the user manually controls paging and adjusts the number of images per page, depending on the density of relevant shots.

In this demo we present a novel approach for labeling clusters in minimally annotated data archives. We propose to build on clustering by aggregating the automatically tagged semantics. We propose and compare four techniques for labeling the clusters and evaluate the performance compared to human labeled ground-truth. We define the error measures to quantify the results, and present examples of the cluster labeling results obtained on the BBC stock shots and broadcast news videos from the TRECVID-2005 video data set.

Without textual descriptions or label information of images, searching semantic concepts in image databases is a very challenging task. Automatic annotation techniques for images are aimed to detect objects which are located visually inside images, like a tiger in grass. One challenge which remains to be solved is “Understanding”. Image understanding is something beyond automatic annotation scope. The second issue is manual annotation of images. In manual annotation, user contribution is important. In this demo, we have developed an online tool which simulates a collaborative environment to help users to generate several facts for selected images. Ontological thinking led us to devise a method for a simple user interface. We are also studying the construction of synergies out of generated facts for our future work.

Semantic search and retrieval of multimedia content is a challenging research field that has drawn significant attention in the multimedia research community. With the dramatic growth of digital media at home, in enterprises, and on the web, methods for effective indexing and search of visual content are vital in unlocking the value of this content. Conventional database search and text search over large textual corpora are both well-understood problems with ubiquitous applications. However, search in non-textual unstructured content, such as image and video data, is not nearly as mature or effective. A common approach for video retrieval, for example, is to apply conventional text search techniques to the associated closed caption or speech transcript. This approach works fairly well for retrieving named entities, such as specific people, objects, or places. However, it does not work well for generic topics related to general settings, events, or people actions, as the speech track rarely describes the background setting or the visual appearance of the subject. Text-based search is not even applicable to scenarios that do not have speech transcripts or other textual metadata for indexing purposes (e.g., consumer photo collections). In addition, speech-based video retrieval frequently leads to false matches of segments that talk about but do not depict the entity of interest. Because of these and other limitations, it is now apparent that conventional text search techniques on their own are not sufficient for effective image and video retrieval, and they need to be combined with techniques that consider the visual semantics of the content. The most substantial work in this field is presented in the TREC Video Retrieval Evaluation (TRECVID) community, which focuses its efforts on evaluating video retrieval approaches by providing common video datasets and a standard set of queries.