Abstract: Provided is an apparatus for encoding an image that adaptively filters reference pixels according to an intra prediction mode and a size of a prediction block, and generates the prediction block according to the intra prediction mode. When the reference pixels for a second directional intra prediction mode are filtered, the reference pixels for a first directional intra prediction mode, that is closer to an intra prediction mode having a direction of 45° with reference to a horizontal mode than the second directional intra prediction mode is, are also filtered. The first and the second intra prediction modes exist between the horizontal mode and the intra prediction mode having the direction of 45° with reference to the horizontal mode. The plurality of sub-blocks and non-zero coefficients of each sub-block are scanned according to a scan pattern determined by the intra prediction mode of the prediction block.

Abstract: A generic categorization method may include receiving interaction data on a distributed computing system operating on a plurality of computing nodes. The distributed computing system may distribute the received interaction data across the plurality of nodes. On each node, categorization rules may be applied to the interaction data via parallel processing. The results, which may include a category associated with each interaction, may be written to a distributed storage system. A user interface may allow a user to define the categorization rules and schemas of interaction data.

Abstract: Methods and apparatus to parallelize data decompression are disclosed. A method selects the initial starting positions in a compressed data bitstream. A first one of the initial starting positions is adjusted to determine a first adjusted starting position by decoding the bitstream starting at a training position in the bitstream. The decoding includes traversing the bitstream from the training position as though first data located at the training position is a valid token. The first decoded data generated by decoding a first segment of the bitstream starting from the first adjusted starting position is output. The first decoded data is merged with second decoded data generated by decoding a second segment of the bitstream. The decoding of the second segment starting from a second position in the bitstream is performed in parallel with the decoding of the first segment. The second segment precedes the first segment in the bitstream.

Abstract: A method for encoding a frame having a plurality of blocks in a video stream using transform-domain intra prediction is disclosed. The method includes generating, using a two-dimensional transform, a set of transform coefficients for a current block; generating, using a one-dimensional transform, a set of transform coefficients for previously coded pixel values in the frame; determining, using the set of transform coefficients for the previously coded pixel values, a set of transform coefficients for a prediction block; determining a residual based on the difference between the set of transform coefficients for the current block and the set of coefficients for the prediction block; and encoding the residual.

Abstract: An operating system uses non-bit aligned test masks to encode compound logical tests within the test mask. Generally, a bit within the test mask will indicate whether the test mask is a bit-aligned test mask or a non-bit-aligned test mask. If the system detects that the test mask in a non-bit-aligned test mask, the system will traverse the test mask to extract bit-aligned sub-masks and perform multi-level logical tests with the bit-aligned sub-masks. Such a system is particularly useful when performing a compound AND-OR logical test involving mutually exclusive event group flags.

Abstract: Provided are a motion estimation method and a motion estimation apparatus. The motion estimation apparatus includes a first register storing information on whether to detect first detection positions, a second register storing information on distances and number information of valid distance information, a controller receiving a command, a shifter, in response to the shift-enable signal, shifting and outputting reference data in a detection region of a reference frame and outputting the received reference data as it is, a selector, in response to the selection signal, selecting and outputting a part of output data of the shifter or outputting the whole output data, a process element (PE) array receiving current data of a current frame, and a comparator generating operation results for respective block sizes using operation results of the plurality of the PEGs.

Abstract: Systems and methods for compressing a raw logical data object (201) for storage in a storage device operable with at least one storage protocol, creating, reading, writing, optimizatic in and restoring thereof. Compressing the raw logical data object (201) comprises creating in the storage device a compressed logical data object (203) comprising a header (204) and one or more allocated compressed sections with predefined size (205-1-205-2); compressing one or more sequentially obtained chunks of raw data (202-1-202-6) corresponding to the raw logical data object (201) thus giving rise to the compressed data chunks (207-1-207-6); and sequentially accommodating the processed data chunks into: said compressed sections (205-1-205-2) in accordance with an order said chunks received, wherein said compressed sections serve as atomic elements of compression/decompression operations during input/output transactions on the logical data object.

Abstract: A circuit for identifying one or more bit positions of instances of a selected bit value in an N-bit input bit string includes a plurality of adders that compute, in parallel, sums of bits in each of P input substrings comprising the input bit string. A plurality of zero position detectors detect, for each of the P input substrings for which a corresponding sum differs from a threshold sum, one or more bit positions of the selected bit value. Correction logic generates adjustment indications indicative of a number of detected instances of the selected bit value. A plurality of output substring adjusters that, based on the detected bit positions and the adjustment indications, collectively output one or more output vectors identifying a bit position of at least an Mth instance of the selected bit value in the input bit string.

Abstract: Systems and methods for transforming a logical data object for storage in a storage device operable with at least one storage protocol, creating, reading, writing, optimization and restoring thereof. Transforming the logical data object comprises creating in the storage device a transformed logical data object comprising one or more allocated storage sections with a predefined size; transforming one or more sequentially obtained chunks of obtained data corresponding to the transforming logical data object; and sequentially storing the processed data chunks into said storage sections in accordance with a receive order of said chunks, wherein said storage sections serve as atomic elements of transformation/de-transformation operations during input/output transactions on the logical data object.

Abstract: A main data input and a lookahead input are held in a holding register. Consecutive overlapping portions of the main data input and the lookahead input are provided to a plurality, M, of half-decoders, which include a subset of frequently-occurring code words of a Huffman code. When no code word not available in the half-decoders is encountered, the half-decoders decode, in parallel, in a single clock cycle, M of the frequently-occurring code words. When a code word not available in the half-decoders is encountered, input intended for a corresponding one of the half-decoders, which input includes the code word not available in the corresponding one of the half-decoders, is applied to an input of a full decoder implemented in ternary content-addressable memory. The full decoder includes all code words of the Huffman code.

Abstract: For boosting decompression in the presence of reoccurring Huffman trees, a previous Huffman tree is used to decompress a new data block if a match is identified between a compacted description of the Huffman tree and a recently used Huffman tree compaction.

Abstract: A decoder level shifter device includes a first decoder level shifter and a second decoder level shifter. The first decoder level shifter has first to fourth input terminals, first and second output terminals, first and second enable terminals, first and second reset terminals. The first to fourth input terminals receive first to second input signals and their complementary signals, respectively. The second decoder level shifter has fifth to eighth input terminals, third to fourth output terminals, and third and fourth enable terminals. The fifth to eighth input terminals receive the first and second input signals and their complementary signals, respectively. The first, second, third, and fourth enable terminals are connected to the fourth, third, second, and first output terminals, respectively.

Abstract: A method of searching for sets of data in a very large columnar database is described.

Abstract: For boosting decompressionin the presence of reoccurring Huffman trees, a compacted description of a Huffman tree is compared to recently used Huffman tree compactions for identifying matches rather than reconstructing a Huffman tree from a header of a compressed block for each data block to avoid overhead of a Huffman tree reconstruction.

Abstract: Various embodiments provide lossless compression of an enumeration space for genetic founder lines. In one embodiment, an input comprising a set of genetic founder lines and a maximum number of generations G is obtained. A set of genetic crossing templates of a height h is generated. A determination is made if at least a first genetic crossing template in the set of genetic crossing templates is redundant with respect to a second genetic crossing template in the set of genetic crossing templates. Based on the at least first genetic crossing template being redundant is redundant with respect to the second genetic crossing template, the at least first genetic crossing template is removed from the set of genetic crossing templates. This process of removing the at least first genetic crossing template from the set of genetic crossing templates the redundant creates an updated set of genetic crossing templates.

Abstract: For boosting decompression in the presence of reoccurring Huffman trees, a compacted description of a Huffman tree is compared to recently used Huffman tree compactions for identifying matches rather than reconstructing a Huffman tree from a header of a compressed block for each data block to avoid overhead of a Huffman tree reconstruction.

Abstract: A decompression engine may include an input configured to receive an input code comprises one or more bits from a bitstream of encoded data, a symbol decoder coupled with the input, where the symbol decoder is configured to calculate, based on the input code, a plurality of candidate addresses each corresponding to a code group. The symbol decoder may further include a group identifier module coupled with the symbol decoder, wherein the group identifier module is configured to identify one of the plurality of code groups corresponding to the input code, and a multiplexer coupled with the group identifier module, wherein the multiplexer is configured to select as a final address one of the plurality of candidate addresses corresponding to the identified code group.

Abstract: According to one embodiment, method for decoding encoded data comprises a hardware module including circuitry to process a data stream. The data stream includes one or more encoded symbols each including a code assigned to a corresponding symbol. A set of least frequently occurring symbols are assigned a common code to encode those symbols within the data stream. Data blocks are generated each containing a data stream portion. One or more encoded symbols within each data block are identified by comparing data block portions to maximum code values for corresponding code lengths to determine a code length for each data block portion. A starting location for the identified encoded symbols within each data block is determined based on the determined code lengths.

Abstract: A binary allocation in a hierarchical coding/decoding comprising a coding/decoding of a digital signal enhancement layer. The signal comprises a succession of L samples, each sample being represented by a mantissa and an exponent. The method comprises the allocation of a predetermined number Nb of enhancement bits to a part at least of the L samples of highest exponent values. In particular, the method comprises the steps: a) enumerating the exponents of the L samples each having a given value, b) calculating at least one aggregate of enumerations of exponents by decreasing values of exponent until the predetermined number Nb is approximated from above, for c) determining a threshold value of largest exponent iexp0 of sample for which no more enhancement bit is available, and allocating the Nb enhancement bits, according to chosen rules, to the samples whose exponent is greater than the aforesaid threshold value iexpo.

Abstract: A method of decoding Huffman-encoded data may comprise receiving a symbol associated with the Huffman encoded data, selecting a target group for the symbol based on a bit length value associated with the symbol, associating the symbol with the target group, associating the symbol with a code, and incrementing a starting code for each of a plurality of groups associated with a starting code that is equal to or greater than the starting code of the target group.

Abstract: An entropy decoder and method for decoding code words with an indication of associated probability for each code word. The decoder can be operable to receive code words and the indication of associated probability. A branch node block can be in communication with a leaf node block. The branch node block comprising one or more branch node lookup tables and branch node control logic. The branch node control logic operable to process a code word in the input buffer using a selected table from the one or more branch node lookup tables to obtain leaf node information and a bit count of a code word size, the branch control logic further operable to refresh the input buffer to replace the bit count of the code word size and to make the leaf node information and the table selection available to the leaf node block.

Abstract: A data predicted value generating unit generates a predicted value (data predicted value) for original data intended to be encoded, based on a history of original data which is floating-point data. A data predicted value modifying unit adjusts a mantissa value of the data predicted value by aligning an exponent value of the data predicted value with an exponent value of the original data. A first residual generating unit generates a residual (first residual) between new original data and the data predicted value after being adjusted. A first residual predicted value generating unit generates a predicted value for the first residual (first residual predicted value), based on a history of first residuals. A second residual generating unit generates a residual (second residual) between the first residual and the first residual predicted value. A residual encoding unit generates encoded data by encoding the second residual.

Abstract: A system, method and computer program product having optimal matching to a known or measured probability distribution encodes data without the use of an excessively large lookup table. An encoder constructed according to the present invention uses two or more different encoding methods in combination. In one embodiment, Huffman coding by table lookup is combined with computational generation, such as by using an exponential Golomb equation. The most commonly occurring elements are looked up in a small Huffman table, while the remaining elements are coded with the equation. In another embodiment, data is encoded using two or more equations. In yet another embodiment, data is encoded using multiple tables in conjunction with one or more equations.

Abstract: A method for decoding encoded data includes receiving data encoded by replacing each of a plurality of characters with bit strings. The method also includes recording, on the basis of definition information, at least one of the characters as corresponding to each of the bit lengths, and generating decode information based on the number of characters, wherein the decode information includes bit string information for sorting the bit strings in a bit length order that is a predetermined order associated with bit lengths. The method also includes, in response to receiving a particular bit length, generating character information in which the characters are sorted in the bit length order by inserting a character corresponding to the particular bit length into a position corresponding to the particular bit length in an array in which at least one of the bit lengths.

Abstract: According to one embodiment, a data processing system has an encoder and a decoder. The encoder is configured to variable-length encode input data to generate an encoded stream. The decoder is configured to decode the encoded stream to generate output data. The encoder has a variable length encoder, a code converter, and a buffer. The variable length encoder is configured to variable-length-encode the input data to generate first variable length codes. The code converter is configured to convert n first variable length codes into a second variable length code. The buffer is configured to buffer the second variable length code to generate the encoded stream.

Abstract: Data compression using a combination, of content independent data compression and content dependent data compression. In one aspect, a method for compressing data comprises: determining whether or not a parameter or attribute of data within a data block is identified for the data block wherein the determining is not based solely on a descriptor that is indicative of the parameter or attribute of the data within the data block; and compressing the data block with at least one encoder associated with the parameter or attribute of the data within the data block to provide a compressed data block.

Abstract: A first encoding unit encodes, within the compression target data, each part with a sequence of the same value into the number of consecutive runs of the value. A decomposing unit decomposes the number of consecutive runs of the same value into an addition of integers belonging to a predetermined integer group. A calculating unit calculates the probability of occurrence of each integer obtained by the decomposition. A second encoding unit encodes each integer by assigning shorter codes to integers with higher probabilities of occurrence.

Abstract: A method and apparatus for symbol-space based compression of patterns are provided. The method comprises receiving an input sequence, the input sequence being of a first length and comprising a plurality of symbols; extracting all common patterns within the input sequence, wherein a common pattern includes at least two symbols; generating an output sequence responsive of the extraction of all common patterns, wherein the output sequence has a second length that is shorter than the first length; and storing in a memory the output sequence as a data layer, wherein the output sequence is provided as a new input sequence for a subsequent generation of a data layer.

Abstract: According to one embodiment, a computer-implemented method of decompressing compressed data is described. A first decompression dictionary is analyzed, the first decompression dictionary including a plurality of chains each with uncompressed data portions distributed in a non-contiguous manner within the first decompression dictionary based on an addressing scheme, where the uncompressed data portions of each chain form a corresponding uncompressed version of compressed data. A second decompression dictionary is generated by combining the uncompressed data portions of each of the chains in the first decompression dictionary to form uncompressed versions of compressed data and instructions are inserted within the second decompression dictionary to decompress compressed data. The compressed data are decompressed by applying the compressed data to the second decompression dictionary.

Abstract: A method of decoding Huffman-encoded data may comprise receiving a symbol associated with the Huffman encoded data, selecting a target group for the symbol based on a bit length value associated with the symbol, associating the symbol with the target group, associating the symbol with a code, and incrementing a starting code for each of a plurality of groups associated with a starting code that is equal to or greater than the starting code of the target group.

Abstract: Variable length code decoding device for decoding variable length code data, including: decoding process tables each including at least two kinds of formats consisting a first format storing identification information for designating a subsequent table to be referred to in a subsequent decoding process, and a second format that stores a decoded value obtained by repeating the decoding process and a significant bit length to be referred to with respect to variable length code data. The device utilizes first, second, third and fourth formats and relative addresses.

Abstract: A compression method applies a selection rule to input symbols and generates a reduced partial set of symbols. The partial set is checked against a dictionary-index for a match. A match identifies a range of matching symbols in a dictionary. The length of the matching range is iteratively increased by checking previous and next symbols in the input data and the dictionary until a matching range length meets a threshold limit or the length of the matching range cannot be increased further. Compressed data corresponding to the input symbols is provided where input symbols are copied over and symbols in a matched range of data are replaced with a representation of their corresponding start location and length in the dictionary.

Abstract: A recording medium stores an information processing program that causes a computer to execute storing a compression symbol map group having a bit string indicating for each character code, presence or absence of the character code in a file group, and a Huffman tree whose leaf corresponding to the character code has a pointer to a compression symbol map of the character code, the Huffman tree converting the character code into a compression symbol of the character code; compressing sequentially and according to the Huffman tree, a character code to be compressed and described in a file of the file group; detecting access to the leaf at the compressing; identifying by a pointer in the accessed leaf, a compression symbol map of the character code to be compressed; and updating a bit that indicates presence or absence of the character code to be compressed, in the identified compression symbol map.

Abstract: According to one embodiment, a computer-implemented method of decompressing compressed data is described. A first decompression dictionary is analyzed, the first decompression dictionary including a plurality of chains each with uncompressed data portions distributed in a non-contiguous manner within the first decompression dictionary based on an addressing scheme, where the uncompressed data portions of each chain form a corresponding uncompressed version of compressed data. A second decompression dictionary is generated by combining the uncompressed data portions of each of the chains in the first decompression dictionary to form uncompressed versions of compressed data and instructions are inserted within the second decompression dictionary to decompress compressed data. The compressed data are decompressed by applying the compressed data to the second decompression dictionary.

Abstract: A hybrid mechanism whereby hardware acceleration is combined with software such that the compression rate achieved is significantly increased while maintaining the original compression ratio (e.g., using full DHT and not SHT or an approximation). The compression acceleration mechanism is applicable to a hardware accelerator tightly coupled with the general purpose processor. The compression task is divided and parallelized between hardware and software wherein each compression task is split into two acceleration requests: a first request that performs SHT encoding using hardware acceleration and provides post-LZ frequency statistics; and a second request that performs SHT decoding and DHT encoding using the DHT generated in software.

Abstract: A parallel test payload includes a bit sequence configured to be segmented into a plurality of sub-sequences having variable bit length carriers. Respective carriers are represented uniformly in each one of the plurality of sub-sequences.

Abstract: A method for data transmission from a programming device to a control unit, the data being compressed by Huffman coding and transmitted with a Huffman coding tree for decoding from the programming device to the control unit, are decompressed and stored in the control unit, by performing, for a predetermined number of data words of N-bits, a first Huffman coding with a first coding tree; performing, for the remaining data words consisting of M-bits, a second Huffman coding with a second coding tree, N greater than M, the predetermined number of data words of N-bits and the remaining data words of M-bits consisting of more than two data words, generating coding trees for the N-bit and the M-bit data words, and transmitting them to the control unit, and providing all coded N-bit data words and N/M of successive M-bit data words with a preceding code bit.

Abstract: A system and method combine an entropy coding algorithm like Huffman or arithmetic coding with a fixed length coding scheme and can improve the compression performance in a fixed rate compression scheme. The fixed length code assigns codes with a fixed length to symbols that have the highest probability of occurrence. Therefore, fixed length coding is used if all symbols in the sequence are from the set of symbols that have the highest probability values. Otherwise an entropy coding algorithm (e.g. Huffman coding) is used to encode quantized symbols.

Abstract: An entropy decoding apparatus may include a data structure stored in memory. The data structure may include a decoding engine vector or context engine vector. The decoding engine vector many have a first set of bits representing a value corresponding to a state of a coding engine, a second set of bits representing an offset value, and a third set of bits representing the contents of an input stream buffer. The context vector may have a first set of bits representing an addresses of a context most probable state, a second set of bits representing a plurality of possible values corresponding to a least probable symbol state of a coding engine, a third set of bits representing an addresses of a context least probable state, a fourth set of bits representing a binary most probable symbol value, and a fifth set of bits representing a binary least probable symbol value.

Abstract: Systems and methods for providing fast and efficient data compression using a combination of content independent data compression and content dependent data compression. In one aspect, a method for compressing data comprises the steps of: analyzing a data block of an input data stream to identify a data type of the data block, the input data stream comprising a plurality of disparate data types; performing content dependent data compression on the data block, if the data type of the data block is identified; performing content independent data compression on the data block, if the data type of the data block is not identified.

Abstract: A mechanism is provided in a data processing system for pipelined compression of multi-byte frames. The mechanism combines a current cycle of data in an input data stream with at least a portion of a next cycle of data in the input data stream to form a frame of data. The mechanism identifies a plurality of matches in a plurality of dictionary memories. Each match matches a portion of a given substring in the frame of data. The mechanism identifies a subset of matches from the plurality of matches that provides a best coverage of the current cycle of data. The mechanism encodes the frame of data into an encoded output data stream.

Abstract: Disclosed herein is a method for decoding a video signal having at least one frame with a plurality of blocks including a current block, including generating, for at least a selected pixel in the current block, a predicted value for at least one pixel located in a row i and a column j of the current block using a processor and according to the following equation: Xij=Li+Aj?C; wherein Xij is the predicted value, Li is the value of a pixel to the left of the current block, Aj is the value of a pixel in a row above the current block and C is the value of a pixel in the row above and the column to the left of the current block and decoding the current block using the predicted value.

Abstract: A method for arithmetic coding of symbols in a stream is described. The method comprises the following steps: coding a current symbol with a current probability model, and updating the current probability model according to the coding of the current symbol. selecting the current probability model in a set of at least two probability models according to a coding cost criterion, and coding an identifier of the selected probability model.

Abstract: A decoding apparatus and method store at least one table including at least one code, receive at least one instruction signal, and extract a symbol value and a symbol length from the at least one table based on the at least one instruction signal. The decoding apparatus calculates a target suffix length that minimizes the size of a generated table and minimizes the size of a non-prefix length of the at least one code.

Abstract: An audio decoder has an arithmetic decoder for providing decoded spectral values on the basis of an arithmetically-encoded representation and a frequency-domain-to-time-domain converter for providing a time-domain audio representation. The arithmetic decoder selects a mapping rule describing a mapping of a code value onto a symbol code in dependence on a numeric current context value describing a current context state. The arithmetic decoder determines the numeric current context value in dependence on a plurality of previously decoded spectral values. The arithmetic decoder evaluates at least one table using an iterative interval size reduction to determine whether the numeric current context value is identical to a table context value described by an entry of the table or lies within an interval described by entries of the table, and derives a mapping rule index value describing a selected mapping table. An audio encoder also uses an iterative interval table size reduction.

Abstract: Systems and methods for optimizing the compression of data using a dictionary coding compression system are described. A system performs an initial compression operation, such as by using a dictionary or other coder. The system then refines the compression by performing an iterative backward reference search for an optimized compressed file.

Abstract: Systems and methods for providing fast and efficient data compression using a combination of content independent data compression and content dependent data compression. In one aspect, a method for compressing data comprises the steps of: analyzing a data block of an input data stream to identify a data type of the data block, the input data stream comprising a plurality of disparate data types; performing content dependent data compression on the data block, if the data type of the data block is identified; performing content independent data compression on the data block, if the data type of the data block is not identified.

Abstract: A method for compressing a data stream includes transforming a data stream into a transformed data stream of referencing symbols and other data elements, the referencing symbols representing a data sequence identical to a data sequence in a reference data block; and encoding the referencing symbols by replacing them with codewords according to an encoding scheme, the transformed stream includes at least one control symbol indicating a change between a portion of the transformed data stream containing a sequence of the other data elements and a portion of the transformed data stream containing a sequence of the codewords for the referencing symbols, the location of the control symbol within the transformed data stream defines the end of the respective portion of the transformed data stream, the encoding scheme providing at least one codeword associated to one of the referencing symbols is longer than a codeword representing the control symbol.

Abstract: A prediction error calculation part calculates a prediction error for each input data. A prediction error encoding part generates a prediction error code by encoding the value of the prediction error. A run-length counting part counts the run-length of the prediction error. When the value of the prediction error changes, a run-length encoding part generates a run-length code by encoding the run-length counted. A code connecting part generates a connected code by connecting the run-length code to the prediction error code of a corresponding prediction error. When the value of the prediction error is a particular value, a prediction error checking part selects a connected code for the prediction error, as an output code. When the value of the prediction error is a different value, the prediction error checking part selects a prediction error code for the prediction error, as an output code. A code output part outputs the output code selected.

Abstract: A size comparison unit reports, to an access control unit, the size of a piece of compressed data having the smallest size from among pieces of compressed data generated via compression processing units compressing blocks, an algorithm name representing a compression processing unit that has generated the piece of compressed data having the smallest size; the access control unit selects a piece of compressed data to be written in a tape medium and reports, to a statistical information processing unit, an algorithm name representing a compression processing unit that has generated this selected piece of compressed data; the statistical information processing unit extracts, from the reported algorithm name, a regularity present in a result of the selection of compressed data and, in accordance with this extracted regularity, specifies and stops a compression processing unit to be stopped.