South Korea
South Korea

Time filter

Source Type

A method of predicting data, representative of an image, for a block X, the method based on a DC coefficient, DC_(A), associated with a block A to the left of the block X, a DC coefficient, DC_(B), associated with a block B above and to the left of the block X, and a DC coefficient, DC_(C), associated with a block C above the block X, the method comprising:if |DC_(A)-DC_(B)|<|DC_(B)-DC_(C)|, then, predicting the data for the block X based on the block C, otherwise, predicting the data for the block X based on the block A.


Patent
AT & T Corporation | Date: 2011-01-05

A method of coding an image that includes a background image and a foreground image occupying respective image areas, the method comprising:coding the background image as a plurality of image coefficients;canceling any image coefficients associated with spatial areas occupied by the foreground image to yield remaining coefficients;reconstructing image data in the respective image areas from the remaining coefficients to yield reconstructed image data; andsubstituting background image data for the reconstructed image data in a background image area yielding output coefficients and, thereafter, outputting the output coefficients as coded image data.


A method and apparatus for increasing the data rate and providing antenna diversity using multiple transmit antennas is disclosed. A set of bits of a digital signal are used to generate a codeword. Codewords are provided according to a channel code. Delay elements may be provided in antenna output channels, or with suitable code construction delay may be omitted. n signals represent n symbols of a codeword are transmitted with n different transmit antennas. At the receiver MLSE or other decoding is used to decode the noisy received sequence. The parallel transmission and channel coding enables an increase the data rate over previous techniques, and recovery even under fading conditions. The channel coding may be concatenated with error correction codes under appropriate conditions.


Patent
AT & T Corporation | Date: 2012-10-10

A method comprising: receiving a request from a subscriber to translate a video signal from a source language to a target language, wherein the video signal comprises text information in the source language; charging the subscriber a fee for on-demand translation during a paid-for period, after which the on-demand translation expires; generating translated information in the target language based on the text information in the source language during the paid-for period; and outputting the video signal embedded with the translated information.


Patent
AT & T Corporation | Date: 2011-02-23

The present invention relates to a method of decoding a bitstream, the method comprising: (a) determining a quantization parameter from quantization data in the bitstream; (b) generating a chrominance scalar that: (i) equals 8 if the quantization parameter has any of values 1 through 4; (ii) equals (the quantization parameter +13)/2 if the quantization parameter has any of values 5 through 24, inclusive of the values 5 and 24; and (iii) equals the quantization parameter minus 6 if the quantization parameter has any of values 25 through 31; (c) inverse quantizing, using the chrominance scalar, a DC coefficient of a chrominance block to yield an inverse quantized DC coefficient; and (d) transforming data of the chrominance block, including the inverse quantized DC coefficient, according to an inverse discrete cosine transform.


Patent
AT & T Corporation | Date: 2011-02-23

The present invention relates to a method of decoding a bitstream, the method comprising: (a) determining a quantization parameter from quantization data in the bitstream; (b) generating a luminance scalar that: (i) equals 8 if the quantization parameter has any of values 1 through 4; (ii) equals 2 times the quantization parameter if the quantization parameter has any of values 5 through 8; (iii) equals the quantization parameter plus 8 if the quantization parameter has any of values 9 through 24; and (iv) equals 2 times the quantization parameter minus 16 if the quantization parameter has any of values 25 through 31; (c) inverse quantizing, using the luminance scalar, a DC coefficient of a luminance block to yield an inverse quantized DC coefficient; and (d) transforming data of the luminance block, including at least the inverse quantized DC coefficient, according to an inverse discrete cosine transform.


A method for use in an encoder, the method comprising:generating a bit vector, for which bit vector a respective position is associated with a corresponding transform coefficient according to an ordered scan of a block of transform coefficients, wherein the respective position in the bit vector has a 1 to indicate a non-zero transform coefficient in a corresponding position in the ordered scan and a 0 to indicate a zero transform coefficient in a corresponding position in the ordered scan; andtransmitting the bit vector away from the encoder.


A method and apparatus for increasing the data rate and providing antenna diversity using multiple transmit antennas is disclosed. A set of bits of a digital signal are used to generate a codeword. Codewords are provided according to a channel code. Delay elements may be provided in antenna output channels, or with suitable code construction delay may be omitted. n signals represent n symbols of a codeword are transmitted with n different transmit antennas. At the receiver MLSE or other decoding is used to decode the noisy received sequence. The parallel transmission and channel coding enables an increase the data rate over previous techniques, and recovery even under fading conditions. The channel coding may be concatenated with error correction codes under appropriate conditions.


A coding strategy efficiently codes intra (macroblocks, regions, pictures, VOP) data. This strategy uses two basis approaches, a Fixed Deinterleaved Transform Coding approach, and an Adaptive Deinterleaved Transform Coding approach. Furthermore, within each approach, two types of coders are developed. One coder operates on an entire picture or VOPs and the other coder operates on small local regions. Using coders and decoders of the present invention, efficient coding at a range of complexities becomes possible, allowing suitable tradeoffs for a variety of applications.


A method for computing a prediction value, for a block X representative of image data, the method comprising: determining whether any of the following are outside of a boundary of a video object plane: a block A, horizontally adjacent to and left of the block X; a block B, vertically adjacent to and above the block A; and a block C, vertically adjacent to and above the block X; setting a DC coefficient for any of the block A, the block B and the block C that is outside the boundary of the video object plane; identifying a DC coefficient for any of the block A, the block B and the block C that is not outside the boundary of the video object plane; wherein the setting and identifying result in one DC coefficient for the block A, DC_(A), one DC coefficient for the block B, DC_(B), and one DC coefficient for the block C, DC_(C); and computing, after the setting and the identifying, a prediction value for the block X, wherein the only DC coefficients that are used in the computing are the DC_(A), the DC_(B), and the DC_(C).

Loading ATT Co. collaborators
Loading ATT Co. collaborators