Abstract: Disclosed is a computer implemented system that facilitates optimized panoramic stitching, comprising an interface that receives images corresponding to different camera locations. Each image comprises a first image depicting a portion of a scene from the point of view of a first camera location and a second image depicting a portion of a scene from the point of view of a second camera location.

Abstract: In a teleconferencing system, multiple parties participate in a teleconference from different locations. The system is configured such that a filmstrip containing simultaneous images or icons representing the participants are displayed on one or more display screens at a conference location in a spatial relationship that represents their real-world or virtual-world spatial relationships.

Abstract: Video data that is associated with a videoconference or a telepresence session is captured and a video signal relating thereto is provided. The video signal is divided into segments, which are identical from one video frame to another subsequent video frame within the video signal. Each segment is encoded independently, to result in encoded segment data such that both I-frames and P-frames are generated for a given segment, so as to support independent reconstruction of segments at a distant end. The encoded segment data is transmitted subsequently to the distant end.

Abstract: Visual information over a network by filtering a full resolution image signal to produce a series of sub-bands, wherein at least one of the sub-bands constitutes a base sub-band representing an image of reduced resolution and wherein additional sub-bands carry higher resolution information. At least the base sub-band is transmitted to a far end endpoint to permit image reconstruction at the far end endpoint at a degree of resolution dependent on the number of sub-bands used to reconstruct the image.

Abstract: A digital all-pass filter has an input port leading to an input sum block and a first feed forward path. Within the first feed forward path is a multiplier. The filter also has an output port coupled to an output sum block that receives a signal from the first feed forward path. A first feedback path is also provided from the output port to the input sum block. The first feedback path includes a multiplier therein. Nested within this structure is a first order all-pass filter having a feed forward path including a forward path delay and forward path that is delayed and a feedback path absent a separate delay element and beginning after the forward path delay element.

Abstract: A method is disclosed for managing multiple data streams transported over a common communications resource in a packet network, wherein data flowing through the resource travels in both directions, and wherein each stream is subject to data peaks. The round trip delay is determined for each data stream, and the transmission of data peaks in one or more of the data streams is delayed to at least reduce the degree of coincidence in the data peaks of different streams without increasing the maximum round trip delay for the data streams.

Abstract: Exchanging of videoconference data between a first endpoint and a second endpoint via a network includes receiving, at the first endpoint, statistical data relating to transmission of first videoconference data. In particular, the first videoconference data is data that is transmitted previously from the first endpoint to the second endpoint via the network, and the statistical data relates to network performance characteristics during transmission of the first videoconference data. An approximately optimized data transmission rate for the network performance characteristics is determined, based on the statistical data. The output bit rate of a video encoder associated with the first endpoint is adjusted, such that second videoconference data, relating to the same videoconference at a time that is later than the first videoconference data, is encoded to provide output data at an adjusted output bit rate for being transmitted via the network at approximately the optimized data transmission rate.

Abstract: An active stream is selected from one of a plurality of audio streams generated in a common acoustic environment by obtaining, for each stream obtaining, at a series of measurement instants tn, where n=1 . . . N, a final performance metric that is representative of the stream's goodness for representing near-end speech at each measurement instant tn. The final performance metrics for each stream are accumulated to determine an overall performance score. A switch to a new stream as the active stream occurs when the stream with the best overall performance score exceeds the overall performance score of the currently active stream by a threshold amount.

Abstract: Processing requirements for acoustic echo cancellation in voice communications are significant and are even more so as the bandwidth of the communication increases. Whilst voice communication occupies a relatively narrow band of frequencies the processing requirements and so forth for wideband communication render acoustic echo cancellation difficult to achieve in a cost effective manner. The invention provides for acoustic echo cancellation within wideband communications by dividing the communications into sub-bands and applying acoustic echo cancellation to some sub-bands whilst processing other sub-bands according to the status of the communications. Additional sub-bands are transmitted at either full-duplex or half-duplex.

Abstract: Within audio telecommunications for audio-visual conferencing a means of reliable double-talk detection is required for several reasons, including control of residual echo, and control of adaptive algorithms used to estimate the echo. Existing techniques tend to either fail when echo return loss is very low or require highly complex algorithms to increase the range of detecting whether double-talk is occurring. The invention is based upon a new method of estimating the power envelope of the double-talk signal. This estimation is then employed in conjunction with a slightly modified version of an echo return loss enhancement (ERLE) based statistic, to form a double-talk decision based on a combination of the two statistics. The resulting decision rule works for both low-ERLE and high-ERLE situations, which is useful when there can be large variations in the magnitude of the echo path seen by the algorithms, or where near-end noise prevents the achievement of high ERLE.

Abstract: An original image is up-sampled to a final image by constructing at least two sub-banded filtered images of the original image with 2D wavelet-based decomposition filters, each filtered image being of the same resolution as the original image. Each sub-banded filtered image is then mapped into a larger filtered image of the same size as the final image. Pixels in each larger filtered image which were not mapped in from the sub-banded filtered images are interpolated or left blank. The larger filtered images are then filtered with 2D reconstruction filters and combined to form the final up-sampled image. The invention has the advantage that high quality up-sampled images can be created in real time suitable for high definition video.

Abstract: A method is disclosed for managing bandwidth in a packet network carrying time sensitive data wherein sessions are established over virtual connections between respective pairs of endpoints. For at least some of the sessions, data representing local session bandwidth usage is sent over the network to a central server. The global bandwidth usage is monitored at the central server based on the local session bandwidth usage obtained from the data received from the endpoints. The central server sends messages to the endpoints to allocate bandwidth to the sessions in accordance with global bandwidth demands on the network.

Abstract: Visual information over a network by filtering a full resolution image signal to produce a series of sub-bands, wherein at least one of the sub-bands constitutes a base sub-band representing an image of reduced resolution and wherein additional sub-bands carry higher resolution information. At least the base sub-band is transmitted to a far end endpoint to permit image reconstruction at the far end endpoint at a degree of resolution dependent on the number of sub-bands used to reconstruct the image.

Abstract: An active stream is selected from one of a plurality of audio streams generated in a common acoustic environment by obtaining, for each stream obtaining, at a series of measurement instants tn, where n=1 . . . N, a final performance metric that is representative of the stream's goodness for representing near-end speech at each measurement instant tn. The final performance metrics for each stream are accumulated to determine an overall performance score. A switch to a new stream as the active stream occurs when the stream with the best overall performance score exceeds the overall performance score of the currently active stream by a threshold amount.

Abstract: A method is disclosed for managing multiple data streams transported over a common communications resource in a packet network, wherein data flowing through the resource travels in both directions, and wherein each stream is subject to data peaks. The round trip delay is determined for each data stream, and the transmission of data peaks in one or more of the data streams is delayed to at least reduce the degree of coincidence in the data peaks of different streams without increasing the maximum round trip delay for the data streams.

Abstract: Exchanging of videoconference data between a first endpoint and a second endpoint via a network includes receiving, at the first endpoint, statistical data relating to transmission of first videoconference data. In particular, the first videoconference data is data that is transmitted previously from the first endpoint to the second endpoint via the network, and the statistical data relates to network performance characteristics during transmission of the first videoconference data. An approximately optimized data transmission rate for the network performance characteristics is determined, based on the statistical data. The output bit rate of a video encoder associated with the first endpoint is adjusted, such that second videoconference data, relating to the same videoconference at a time that is later than the first videoconference data, is encoded to provide output data at an adjusted output bit rate for being transmitted via the network at approximately the optimized data transmission rate.

Abstract: Processing requirements for echo cancellation in voice communications are significant and are even more so as the bandwidth of the communication increases. Whilst voice communication occupies a relatively narrow band of frequencies the processing requirements and so forth for wideband communication render echo cancellation difficult to achieve in a cost effective manner. The invention provides for provides echo cancellation within wideband communications by dividing the communications into sub-bands and applying echo cancellation to some sub-bands whilst processing other sub-bands according to the status of the communications. Additional sub-bands are transmitted at either full-duplex or half-duplex.

Abstract: Within audio telecommunications for audio-visual conferencing a means of reliable double-talk detection is required for several reasons, including control of residual echo, and control of adaptive algorithms used to estimate the echo. Existing techniques tend to either fail when echo return loss is very low or require highly complex algorithms to increase the range of detecting whether double-talk is occurring. The invention is based upon a new method of estimating the power envelope of the double-talk signal. This estimation is then employed in conjunction with a slightly modified version of an echo return loss enhancement (ERLE) based statistic, to form a double-talk decision based on a combination of the two statistics. The resulting decision rule works for both low-ERLE and high-ERLE situations, which is useful when there can be large variations in the magnitude of the echo path seen by the algorithms, or where near-end noise prevents the achievement of high ERLE.

Abstract: A digital all-pass filter has an input port leading to an input sum block and a first feed forward path. Within the first feed forward path is a multiplier. The filter also has an output port coupled to an output sum block that receives a signal from the first feed forward path. A first feedback path is also provided from the output port to the input sum block. The first feedback path includes a multiplier therein. Nested within this structure is a first order all-pass filter having a feed forward path including a forward path delay and forward path that is delayed and a feedback path absent a separate delay element and beginning after the forward path delay element.

Abstract: Video data that is associated with a videoconference or a telepresence session is captured and a video signal relating thereto is provided. The video signal is divided into segments, which are identical from one video frame to another subsequent video frame within the video signal. Each segment is encoded independently, to result in encoded segment data such that both I-frames and P-frames are generated for a given segment, so as to support independent reconstruction of segments at a distant end. The encoded segment data is transmitted subsequently to the distant end.