Abstract: A method of generating a feature descriptor includes determining a first output histogram of an input by processing a first group of pixels of the input to determine first contributions to bins of the first output histogram. The input image including gradient orientation values and gradient magnitude values of a portion of an image that is in a region of a detected feature. After processing the first group of pixels, the method includes determining a second output histogram of the input by processing a second group of pixels of the input to determine second contributions to bins of the second output histogram.

Abstract: A method of generating a feature descriptor includes determining a first output histogram of an input by processing a first group of pixels of the input to determine first contributions to bins of the first output histogram. The input image including gradient orientation values and gradient magnitude values of a portion of an image that is in a region of a detected feature. After processing the first group of pixels, the method includes determining a second output histogram of the input by processing a second group of pixels of the input to determine second contributions to bins of the second output histogram.

Abstract: A method includes receiving image data and performing a non-maximum suppression (NMS) operation on the image data. The method also includes initiating an edge tracking by hysteresis (ETH) operation on a portion of the image data prior to completion of the NMS operation.

Abstract: Systems and methods pertain to looking up entries of a table. A processor receives one or more single instruction multiple data (SIMD) instructions, including a first SIMD instruction which specifies a first subset of indices. A first subset of table entries is looked up, using a crossbar, with the first subset of indices. A first vector output of the first SIMD instruction is based on whether the outputs of the crossbar belong to a desired subset of table entries. Similarly, second, third, and fourth SIMD instructions specify corresponding second, third, and fourth subsets of indices to lookup the remaining table entries using the crossbar. The size of the crossbar is based on the number of indices in the subset of indices used to lookup table entries.

Abstract: A method includes receiving image data and performing a non-maximum suppression (NMS) operation on the image data. The method also includes initiating an edge tracking by hysteresis (ETH) operation on a portion of the image data prior to completion of the NMS operation.

Abstract: The techniques of this disclosure relate to video encoding and include using an inter mode determination for neighboring blocks, rather than the final prediction mode determination for the neighboring block, when determining an inter mode for a current block. In this way, inter mode and intra mode estimation may be separated and performed in different stages of a multi-threaded parallel video encoding implementation. In addition, this disclosure also proposes generating sub-pixel values in a third stage of the multi-threaded parallel video encoding implementation at a frame level, rather than for each macroblock during inter mode estimation process for that macroblock.

Abstract: In a digital media pipeline, hardware-accelerated transform functions enable longer CPU idle time and a reduction in data transfer between the CPU and hardware, for the primary purpose of conserving power or increasing content security. Multiplexer/de-multiplexer functions can be configured as either stand-alone transform units or as plug-in components to a “pluggable” (host) media source or to a “pluggable” (host) media sink, so that the benefit of hardware acceleration can be applied to the source and sink as well as to the media foundation transform (MFT). Further data processing and control can be routed to a remote processing entity. The disclosed pluggable media source has a single input and one or more outputs; the pluggable media sink has one or more inputs and a single output. The pluggable media source and sink can be configured to accept plug-in components that support a wide range of data formats.

Abstract: Systems and methods for smooth rewind playback of streamed media are provided. The media includes relatively-encoded frames and independently-encoded frames. The method includes receiving a rewind request indicating a rewind speed for rewind playback of the media, selectively dropping relatively-encoded frame(s) based on a receipt constraint and a decoding constraint to form a subset of the media, and receiving frames of the subset. The method further includes selecting, in a reverse order, a selected group of pictures (GOP) included within the subset, and decoding relatively-encoded frame(s) of the GOP in a forward sequential frame order. The method further includes caching relatively-encoded frame(s) of the GOP in the forward sequential frame order, and when caching, dropping and overwriting relatively-encoded frame(s) of the GOP selectively according to a memory constraint and/or a display constraint.

Abstract: Video decoding innovations for using local picture identifiers and computing co-located information are described. In one aspect, a decoder identifies reference pictures in a reference picture list of a temporal direct prediction mode macroblock that match reference pictures used by a co-located macroblock using local picture identifiers. In another aspect, a decoder determines whether reference pictures used by blocks are the same by comparing local picture identifiers during calculation of boundary strength. In yet another aspect, a decoder determines a picture type of a picture and based on the picture type selectively skips or simplifies computation of co-located information for use in reconstructing direct prediction mode macroblocks outside the picture.

Abstract: Video decoding innovations for using local picture identifiers and computing co-located information are described. In one aspect, a decoder identifies reference pictures in a reference picture list of a temporal direct prediction mode macroblock that match reference pictures used by a co-located macroblock using local picture identifiers. In another aspect, a decoder determines whether reference pictures used by blocks are the same by comparing local picture identifiers during calculation of boundary strength. In yet another aspect, a decoder determines a picture type of a picture and based on the picture type selectively skips or simplifies computation of co-located information for use in reconstructing direct prediction mode macroblocks outside the picture.

Abstract: Systems and methods for smooth rewind playback of streamed media are provided. The media includes relatively-encoded frames and independently-encoded frames. The method includes receiving a rewind request indicating a rewind speed for rewind playback of the media, selectively dropping relatively-encoded frame(s) based on a receipt constraint and a decoding constraint to form a subset of the media, and receiving frames of the subset. The method further includes selecting, in a reverse order, a selected group of pictures (GOP) included within the subset, and decoding relatively-encoded frame(s) of the GOP in a forward sequential frame order. The method further includes caching relatively-encoded frame(s) of the GOP in the forward sequential frame order, and when caching, dropping and overwriting relatively-encoded frame(s) of the GOP selectively according to a memory constraint and/or a display constraint.

Abstract: A method and apparatus are provided for mixing a plurality of signals within a predetermined dynamic range without clipping. In the method and apparatus, first and second signal samples are added together to obtain a first intermediate result. Then the first signal sample is multiplied with the second signal sample to obtain a second intermediate result. In one embodiment, the second intermediate result is subtracted from the first intermediate result to obtain a third intermediate result, and the third intermediate result is discarded if the third intermediate result is less than zero. In another embodiment, the second intermediate result is added to the first intermediate result to obtain the third intermediate result, and the third intermediate result is discarded if the third intermediate result is greater than zero. An output signal sample is provided based on the third intermediate result.

Abstract: A method, computer program product, and data processing system for performing automated optimization of a control processing flow containing Boolean conditional expressions is disclosed. Each conditional expression is profiled using a representative set of test vectors to determine the probabilities of individual conditions and combinations of conditions in the expression. Next, the expression is restructured or reordered, including selective computation of subexpressions, based on the probability information, so as to maximize the probability of a short-circuit termination of evaluation of the expression and/or to minimize the statistical execution time of the expression. This process is performed for all Boolean conditional expressions related to the control processing flow.

Abstract: Video decoding innovations for using local picture identifiers and computing co-located information are described. In one aspect, a decoder identifies reference pictures in a reference picture list of a temporal direct prediction mode macroblock that match reference pictures used by a co-located macroblock using local picture identifiers. In another aspect, a decoder determines whether reference pictures used by blocks are the same by comparing local picture identifiers during calculation of boundary strength. In yet another aspect, a decoder determines a picture type of a picture and based on the picture type selectively skips or simplifies computation of co-located information for use in reconstructing direct prediction mode macroblocks outside the picture.

Abstract: A method, computer program product, and data processing system for performing automated optimization of a control processing flow containing Boolean conditional expressions is disclosed. Each conditional expression is profiled using a representative set of test vectors to determine the probabilities of individual conditions and combinations of conditions in the expression. Next, the expression is restructured or reordered, including selective computation of subexpressions, based on the probability information, so as to maximize the probability of a short-circuit termination of evaluation of the expression and/or to minimize the statistical execution time of the expression. This process is performed for all Boolean conditional expressions related to the control processing flow.

Abstract: A method and apparatus are provided for mixing a plurality of signals within a predetermined dynamic range without clipping. In the method and apparatus, first and second signal samples are added together to obtain a first intermediate result. Then the first signal sample is multiplied with the second signal sample to obtain a second intermediate result. In one embodiment, the second intermediate result is subtracted from the first intermediate result to obtain a third intermediate result, and the third intermediate result is discarded if the third intermediate result is less than zero. In another embodiment, the second intermediate result is added to the first intermediate result to obtain the third intermediate result, and the third intermediate result is discarded if the third intermediate result is greater than zero. An output signal sample is provided based on the third intermediate result.