Abstract: A method for assigning servers to provide multiple description bitstreams to a mobile client (in a mobile client environment) or to a fixed client (in a fixed client environment). In one embodiment, the present invention, upon receiving a request from a mobile client to have media data streamed thereto, analyzes a plurality of servers to determine a first candidate server for providing a first multiple description bitstream to the base station along a first path. The present method also determines a second candidate server for providing a second multiple description bitstream to the base station along a second path. The present method then sends a request to the first candidate server to provide the first multiple description bitstream to a mobile client through a base station along the first path, and also sends a request to the second candidate server to provide the second multiple description bitstream to the mobile client through the same base station along a second path.

Abstract: A fixed client and a mobile client for receiving multiple description media streams. In one embodiment, the client comprises a multiple description receiving portion which is adapted to receive a plurality of multiple description bitstreams. The client includes memory coupled to the multiple description receiving portion for storing the plurality of multiple description bitstreams in respective portions thereof. The client of the present embodiment also includes a synchronization module coupled to the memory for blending the plurality of multiple description bitstreams. In one embodiment, a decoder is coupled to the synchronization module for decoding the plurality of multiple description bitstreams. A source control module for determining appropriate operation characteristics of the client is also coupled to the synchronization module. Also, a user interface device is coupled to the decoder to present to a user, media previously encoded into the plurality of multiple description bitstreams.

Abstract: A method and system for securely and scalably encoding data are disclosed. A method and system for decoding data which has been securely and scalably encoding are also disclosed. In one encoding method embodiment, the present invention recites receiving data. The present method then segments the data into corresponding regions. The regions are then scalably encoded into scalable data. The present embodiment then progressively encrypts the scalable data to generate progressively encrypted scalable data. Next, the present embodiment, packetizes the progressively encrypted scalable data.

Abstract: A method and system for transcoding a stream of data packets are disclosed. The data packets are received from a source node and include data that are encrypted. The stream of data packets is operated on by a function that configures the stream according to downstream attributes, such as the attributes of a receiving node or of the communication channel. Significantly, the function is performed without decrypting the encrypted media data. The steam of data packets configured for the downstream attributes is sent to the receiving node. Because the encrypted data do not need to be decrypted and then encrypted again, the computational resources needed for transcoding the stream of data packets is significantly reduced, and the security of the data is not compromised.

Abstract: A computer readable medium having a data packet stored therein for causing a functional change in the operation of a device is disclosed. In one embodiment, the data packet is comprised of a scalably encoded, progressively encrypted data portion. In the present embodiment, the data packet further includes a header data portion corresponding to the scalably encoded, progressively encrypted data portion. The header data portion includes information adapted to be used by a transcoder to transcode the scalably encoded, progressively encrypted, data portion.

Abstract: A predictively-coded object-based picture signal representing a group of pictures is transcoded into a predictively-coded block-based picture signal representing the group of pictures by extracting a coded scene descriptor and coded object descriptors from the predictively-coded object-based picture signal. The coded scene descriptor is decoded to generate a scene descriptor. The coded object descriptors are partially decoded to generate respective partially-decoded object descriptors. The decoding extracts coding information that describes the predictive coding of the coded object descriptors. In response to the scene descriptor, a frame of a partially-encoded block-based picture signal representing one of the pictures as a current picture is generated from the partially-decoded object descriptors.

Abstract: This disclosure provides a reverse play system that may be used with newer digital television and image standards. In particular, a compressed input including motion vectors is received, and the motion vectors are used to calculate closest match positions. The order of image frames is then reversed, and reverse motion vectors are calculated anew. In calculating new motion vectors, compression software performs motion search for each data block by first comparing a local neighborhood around the block with the closest match positions. If there is an exact match (between the data block and a closest match position), then an original motion vector is inverted and exactly re-used as a motion vector in reverse play. If there are several (non-exact) matches, the software weights every motion vector associated with overlap of the neighborhood and selects the motion vector corresponding to the largest overlap.

Abstract: This disclosure provides a block-based compression system which uses a variable size search window, to thereby substantially reduce the time required for compression. In a MPEG-2 system, square tiles are processed left to right in rows, beginning at the top of a frame and finishing at the bottom. The absolute value of pixel differences for already-processed “northwest,” “north,” “northeast” and “west” neighbor tiles are separately summed and compared to a threshold. If any of the sums exceeds the threshold or if any of the neighbor tiles are unavailable (e.g., the tile under consideration is in the top row or the left-most column), then a maximum size search window is used to determine the tile's “closest match” in a prior image frame and the associated motion vector. If none of the exceed the threshold, then a relatively smaller size search window is used.

Abstract: This disclosure provides for an image editing system that permits manipulation of compressed image formats without full decompression to the image domain. In particular, an editor manipulates discrete cosine transform ("DCT") coefficients for a dependent frame in a compressed image sequence to incorporate DCT representations of anchor frames upon which the dependent frame depends. Once converted, the frame can simply be reordered in the compressed domain without violating temporal dependencies, subject to bit rate matching. In one embodiment, an image sequence can be cut to leave remaining image frames which are then altered, so as to eliminate temporal dependencies upon eliminated frames. In another embodiment, an image sequence can be processed in the compressed domain for reverse-play. The editor provided by this disclosure permits real-time compressed domain editing applications, with reduced requirements for computational bandwidth.

Abstract: This disclosure provides a block-based compression system which uses a variable size search window, to thereby substantially reduce the time required for compression. In a MPEG-2 system, square tiles are processed left to right in rows, beginning at the top of a frame and finishing at the bottom. The absolute value of pixel differences for already-processed "northwest," "north," "northeast" and "west" neighbor tiles are separately summed and compared to a threshold. If any of the sums exceeds the threshold or if any of the neighbor tiles are unavailable (e.g., the tile under consideration is in the top row or the left-most column), then a maximum size search window is used to determine the tile's "closest match" in a prior image frame and the associated motion vector. If none of the exceed the threshold, then a relatively smaller size search window is used. The size of the smaller window is selected in dependence upon the motion vectors for the neighbor tiles, e.g.