Abstract: Methods and apparatus are disclosed for automatically selecting a video output from among several video input sources. In one method, one or more audio sensors are associated with each video input source. Preferably, an audio sensor is positioned to receive audio signals from directions that receive favorable coverage in the field of view of the associated video source. An autoselector calculates audio scores for each of the audio sensors over short (e.g., 0.5 seconds) examination intervals. At each examination interval, the potential exists for a different video source to be selected as the video output. The autoselector selects a video source based on the audio scores for an examination interval, as well as the recent time-history of video source selection. For instance, if a new video source has just been selected, selection of a different source may be disabled for a few seconds. The time-history is also used to increase the probability that source selection varies in a seemingly-natural manner.

Abstract: A method of generating a visual program summary is disclosed for a multiple source video feed. In the disclosed method, a computing device continuously captures frames from a set of available video feeds such as television channels. The computing device analyzes the captured video frames to remove redundant frames. The computing device then selects a set of frames to generate a visual program summary. The selected frames are composited together to generate a visual program summary.

Abstract: A method for interfacing with an electronic system is provided. The method includes receiving images from a camera. A portion of at least one image from the camera is tracked. A pattern in the tracked portion of the at least one image is recognized. And, the electronic system is controlled based on the recognized pattern.

Abstract: Fragile watermarking for objects is disclosed. In one embodiment of the invention, a system includes an encoder and a decoder. The encoder encodes a watermark into a object, such as a three-dimensional object, in a fragile manner, utilizing a key. The decoder decodes the watermark from the object in which the watermark is encoded in a pubic manner, also utilizing a key. Visualization techniques for detecting alterations using fragile watermarking are also disclosed.

Abstract: A system and method is provided for using a linearized data structure to display a set of ordered images. The images are ordered based on the attributes of the images. The system and method includes displaying a target image and related images on a graphical user interface, where the related images are near the target image in the linearized data structure. The relative sizes of the target image and the related images may reflect the similarity of the related images to the attributes of the target image.

Abstract: A system and method is provided for creating a linearized data structure for ordering a set of images based on their attributes. The system and method includes characterizing each of a set of images based on the attributes of the images, mapping the set of images onto a curve based on the attributes of the images, and generating an ordered arrangement of the images by traversing the curve. The ordered set of images may be used for displaying the images for efficient viewing and browsing.

Abstract: A system and method for compressing and decompressing multi-spectral images in two color spaces is provided. The system and method employ the transformation, compression and reconstruction of color space components from which difference data between the original color space components and the reconstructed color space components may be derived. A compressed data stream including compressed color space data and compressed difference data in a another color space is generated. The difference data may be utilized to reconstruct a lossless or near lossless image of the original.

Abstract: A system and method for generating bounded-loss color transformations employed in the compression and decompression of multi-spectral images is provided. The bounded-loss color space transformations provide transformations from a first color space to a second color space and back to the first color space with bounded loss or error. The bounded loss or error allows for the near lossless compression and reconstruction of multi-spectral images transformed from a first color space to a second color space and ultimately back to the first color space.

Abstract: A method of displaying images is based on both a first linearization and a second linearization. In one embodiment, the linearizations are performed by traversing two space-filling curves. In another embodiment, the linearizations are performed by traversing a cluster data structure. More than two linearizations may be displayed.

Abstract: A system and method for compressing and decompressing multi-spectral images in two color spaces is provided. The system and method employ the transformation, compression and reconstruction of color space components from which difference data between the original color space components and the reconstructed color space components may be derived. A compressed data stream including compressed color space data and compressed difference data in a another color space is generated. The difference data may be utilized to reconstruct a lossless or near lossless image of the original.

Abstract: According to the invention, channel surfers are aided by summary frames appearing on the screen along with the regular program. These summary frames are embedded in the broadcast and appear in a small window on the screen at some point in the future as the video is broadcast. The summary frames depict key scenes from the past which aid the viewer in quickly ascertaining the current plot or theme of the video program. In an interactive television environment, a short video segment may be associated with each shot such that when the shot is selected, the past video segment is played in a larger window on the screen complete with sound. Once the segment is viewed, the video program is rejoined in progress with the viewer having a better sense of what is going on in the video. Hence, a when a channel surfer arrives at a new channel, rather than only having what is currently playing to catch the eye, summary frames are also available to catch the surfer's attention and aid in understanding the programming.

Abstract: A system and method for generating bounded-loss color transformations employed in the compression and decompression of multi-spectral images is provided. The bounded-loss color space transformations provide transformations from a first color space to a second color space and back to the first color space with bounded loss or error. The bounded loss or error allows for the near lossless compression and reconstruction of multi-spectral images transformed from a first color space to a second color space and ultimately back to the first color space.

Abstract: Given compressed video which supports the coding of interlaced and noninterlaced frames through DCT and motion compensation, shared information is used in a macroblock for efficient conversion of inter-coded blocks to DCT intra-coded blocks. The shared information is also used to efficiently construct spatially reduced images without decompression.

Abstract: Given compressed video which supports the coding of interlaced frames through DCT and motion compensation, different construction methods are applied based on select DCT domain coefficients and/or motion vector informations to produce spatially reduced images without decompressing. The spatially reduced images can then be used for video browsing, authoring and processing purposes.

Abstract: A system and method for compressing and decompressing multi-spectral images in two color spaces is provided. The system and method employ the transformation, compression and reconstruction of color space components from which difference data between the original color space components and the reconstructed color space components may be derived. A compressed data stream including compressed color space data and compressed difference data in a another color space is generated. The difference data may be utilized to reconstruct a lossless or near lossless image of the original.

Abstract: A system and method for browsing dynamic video information over a network includes selecting a small subset of frames from video shots and collections, using a threshold difference between two compared frames, to capture the dynamic content. Further selection and interleaving the selected frames within the shots and collections can be done to satisfy resource constraints like utilization of bandwidth. Interleaving is performed in storage and transmission of the frames to improve presentation of video information and to simultaneously display selected frames on a computer display to convey the dynamic of video. The system and method permit a dynamic summary of video information to be sent to a user over a network, while reducing resources (bandwidth) used by the network for the amount of information presented to the user.

Abstract: A three-dimensional image registration method also uses an algorithm and applies it to the solution of such problems in 3D CT DSA. The method can deal with incomplete or partial volumetric data in registration and correct patient global motion prior to subtraction even when it is coupled with local unconscious/nonrigid movements. Experimental results demonstrate the effectiveness of this algorithm on several clinical spiral CT data.

Abstract: Coding systems that provide a perceptually improved approximation of the short-term characteristics of speech signals compared to typical coding techniques such as linear predictive analysis while maintaining enhanced coding efficiency. The invention advantageously employs a non-linear transformation and/or a spectral warping process to enhance particular short-term spectral characteristic information for respective voiced intervals of a speech signal. The non-linear transformed and/or warped spectral characteristic information is then coded, such as by linear predictive analysis to produce a corresponding coded speech signal. The use of the non-linear transformation and/or spectral warping operation of the particular spectral information advantageously causes more coding resources to be used for those spectral components that contribute greater to the perceptible quality of the corresponding synthesized speech.

Abstract: A new technique for extracting a hierarchical decomposition of a complex video selection for browsing purposes, combines visual and temporal information to capture the important relations within a scene and between scenes in a video, thus allowing the analysis of the underlying story structure with no a priori knowledge of the content. A general model of hierarchical scene transition graph is applied to an implementation for browsing. Video shots are first identified and a collection of key frames is used to represent each video segment. These collections are then classified according to gross visual information. A platform is built on which the video is presented as directed graphs to the user, with each category of video shots represented by a node and each edge denoting a temporal relationship between categories. The analysis and processing of video is carried out directly on the compressed videos.

Abstract: A new technique for extracting a hierarchical decomposition of a complex video selection for browsing purposes, combines visual and temporal information to capture the important relations within a scene and between scenes in a video, thus allowing the analysis of the underlying story structure with no a priori knowledge of the content. A general model of hierarchical scene transition graph is applied to an implementation for browsing. Video shots are first identified and a collection of key frames is used to represent each video segment. These collections are then classified according to gross visual information. A platform is built on which the video is presented as directed graphs to the user, with each category of video shots represented by a node and each edge denoting a temporal relationship between categories. The analysis and processing of video is carried out directly on the compressed videos.