Abstract: A method including receiving an audio signal, generating a transformed audio signal by transforming the audio signal using a plurality of windows each separated in time, generating an interpolated audio signal by interpolating the transformed audio signal, generating a separated audio signal by applying a mask to the interpolated audio signal, and compressing the separated audio signal.

Abstract: A method including receiving a plurality of audio channels based on an audio stream, applying a model based on at least one acoustic perception algorithm to the plurality of audio channels to generate a first modelled audio stream, quantizing the plurality of audio channels using a first set of quantization parameters, dequantizing the quantized plurality of audio channels using the first set of quantization parameters, applying the model based on at least one acoustic perception algorithm to the dequantized plurality of audio channels to generate a second modelled audio stream, comparing the first modelled audio stream and the second modelled audio stream, in response to determining the comparison of the first modelled audio stream and the second modelled audio stream does not meet a criterion, generating a second set of quantization parameters, and quantizing the plurality of audio channels using the second set of quantization parameters.

Abstract: An context-based encoding mechanism uses a predetermined number of bytes in a previous segment of a file to determine a context for the current segment. The current segment is encoded using a dictionary that corresponds to the determined context. An example method includes determining, for a first segment in a data file, a first context state based on a first context segment within the data file that precedes the first segment, identifying a first indexed dictionary from a plurality of indexed dictionaries based on the first context state, and encoding the first segment using the identified first indexed dictionary.

Abstract: A technique for improving progressive encoded JPEG includes displaying an oversmoothed version of an image as the image data is being received. The oversmoothed image may be smoothed according to a smoothing kernel, e.g., a convolution kernel (such as a Gaussian). The oversmoothed image is a first layer over which other image layers are displayed. It is noted that the oversmoothed image may present a recognizable version of the image to a user, including recognizable versions of various image features (e.g., persons, objects). As the other layers are rendered on the display, these image features remain visible to the user. That is, the image features are not artifacts that may disappear with the rendering of final image layers; this may occur with the conventional progressive encoded images and interferes with the user experience.

Abstract: A computer-implemented method can include receiving a first signal corresponding to a first flow of acoustic energy, applying a transform to the received first signal using at least a first amplitude-independent window size at a first frequency and a second amplitude-independent window size at a second frequency, the second amplitude-independent window size improving a temporal response at the second frequency, wherein the second frequency is subject to amplitude reduction due to a resonance phenomenon associated with the first frequency, and storing a first encoded signal, the first encoded signal based on applying the transform to the received first signal.

Abstract: Techniques of color image processing involve performing a transformation for each color channel that mixes intensity values from other channels to produce a new intensity value for that channel. The new intensity values, representing the effect of overlapped response spectra of the S, M, and L cones, then provide values of the sensitivities of the photoreceptors of each of the cones. These values of the sensitivities form the basis of more accurate color image processing. For example, compression ratios of gamma-compressed color images may be increased when more the sensitivities are more accurate.

Abstract: A decoder may perform a method of decompressing images that include texture features that are not aligned with an axis of the image being compressed. In some example implementation, the method may include receiving a block of geometrically transformed pixel values and performing an inverse geometric transformation on the block of geometrically transformed pixel values to generate a first block of pixel values. The geometrically transformed pixel values represent texture features of an image that are non-parallel with a vertical axis or a horizontal axis of the image and the first block of pixel values being one of a plurality of blocks of the image. The example method may further include generating at least a portion of the image based on the first block of pixel values.

Abstract: A method can include compressing a first original frame of a video stream to an intraframe, the intraframe comprising fewer symbols than the first original frame, compressing a second original frame to a first interframe, the first interframe referencing the intraframe and comprising fewer symbols than the second original frame, determining an intraframe error of the intraframe due to the compression of the first original frame, determining a first interframe error of the first interframe due to the compression of the second original frame, determining a compression level for a third original frame based on the intraframe error and the first interframe error, and compressing the third original frame to a second interframe, the second interframe referencing the intraframe and the first interframe and comprising fewer symbols than the third original frame, a number of symbols included in the second interframe being based on the determined compression level.

Abstract: Systems and methods generate reasonably secure hash values at relatively few CPU cycles per byte. An example method includes, for each of a plurality of packets, injecting the packet into an internal state that represents an internal hash sum, mixing the internal state using multiplication, and shuffling the result of the multiplication so that bytes with highest quality are moved to locations that will propagate most widely in a next multiplication operation. Each of the plurality of packets include data from an input to be hashed. In some implementation, a last packet for the input is padded. The method may also include further mixing the internal state using multiplication after processing the plurality of packets and providing, to a requesting process, a portion of the final internal state as a hash of the input.

Abstract: The present disclosure proposes a model that has more expressive power, e.g., can generalize from a smaller amount of parameters and assign more computation in areas of the function that need more computation. In particular, the present disclosure is directed to novel machine learning architectures that use the exponential of an input-dependent matrix as a nonlinearity. The mathematical simplicity of this architecture allows a detailed analysis of its behavior.

Abstract: A computer-implemented method can include receiving a first signal corresponding to a first flow of acoustic energy, applying a transform to the received first signal using at least a first amplitude-independent window size at a first frequency and a second amplitude-independent window size at a second frequency, the second amplitude-independent window size improving a temporal response at the second frequency, wherein the second frequency is subject to amplitude reduction due to a resonance phenomenon associated with the first frequency, and storing a first encoded signal, the first encoded signal based on applying the transform to the received first signal.

Abstract: A method can include compressing a first original frame of a video stream to an intraframe, the intraframe comprising fewer symbols than the first original frame, compressing a second original frame to a first interframe, the first interframe referencing the intraframe and comprising fewer symbols than the second original frame, determining an intraframe error of the intraframe due to the compression of the first original frame, determining a first interframe error of the first interframe due to the compression of the second original frame, determining a compression level for a third original frame based on the intraframe error and the first interframe error, and compressing the third original frame to a second interframe, the second interframe referencing the intraframe and the first interframe and comprising fewer symbols than the third original frame, a number of symbols included in the second interframe being based on the determined compression level.

Abstract: A fast cryptographic hash of an input file using multiplication and permutation operations in a parallel processing environment. An example method includes updating an internal state for each of a plurality of packets, the packets being read from an input file. Updating the state for a packet can include injecting the packet into an internal state, mixing the bits of the internal state using multiplication, and shuffling the result of the multiplication so that bits with highest quality are permuted to locations that will propagate most widely in a next multiplication operation. The method also includes performing a reduction on the internal state and repeating the update of the internal state, the reduction, and the injecting a second time. The method may further include finalizing the internal state and storing a portion of the final internal state as a cryptographic hash of the input file.

Abstract: An encoder and/or a computer implemented encoding method includes a texture module configured to determine texture data associated with texture of an image, a noise module configured to determine noise data based on the texture data, a synthesis module configured to generate spatial spectral characteristics of the noise, and combine at least one of the noise data, the texture data, and the spatial spectral characteristics of the noise based on at least one border between adjacent textures, and an encoding module configured to compress the image using an image compression codec.

Abstract: An improved color space (YHB model) for compressing image files is provided. An example method includes storing a sum of an unweighted first color value and an unweighted second color value for each pixel in a plurality of pixels of an image as a first channel, sub sampling, among the plurality of pixels, a difference between the first color value and the second color value as a second channel, sub sampling, among the plurality of pixels, a third color value as a third channel, and storing the first channel, the second channel, and the third channel as the compressed image. In some implementations, the original image may be split into a low frequency version and a high frequency version. The system may apply the YHB model to the high frequency version and apply a conventional model or a second variation of the YHB model to the low frequency version.

Abstract: A decoder may perform a method of decompressing images that include texture features that are not aligned with an axis of the image being compressed. In some example implementation, the method may include receiving a block of geometrically transformed pixel values and performing an inverse geometric transformation on the block of geometrically transformed pixel values to generate a first block of pixel values. The geometrically transformed pixel values represent texture features of an image that are non-parallel with a vertical axis or a horizontal axis of the image and the first block of pixel values being one of a plurality of blocks of the image. The example method may further include generating at least a portion of the image based on the first block of pixel values.

Abstract: An context-based encoding mechanism uses a predetermined number of bytes in a previous segment of a file to determine a context for the current segment. The current segment is encoded using a dictionary that corresponds to the determined context. An example method includes determining, for a first segment in a data file, a first context state based on a first context segment within the data file that precedes the first segment, identifying a first indexed dictionary from a plurality of indexed dictionaries based on the first context state, and encoding the first segment using the identified first indexed dictionary.

Abstract: Systems and methods generate reasonably secure hash values at relatively few CPU cycles per byte. An example method includes, for each of a plurality of packets, injecting the packet into an internal state that represents an internal hash sum, mixing the internal state using multiplication, and shuffling the result of the multiplication so that bytes with highest quality are moved to locations that will propagate most widely in a next multiplication operation. Each of the plurality of packets include data from an input to be hashed. In some implementation, a last packet for the input is padded. The method may also include further mixing the internal state using multiplication after processing the plurality of packets and providing, to a requesting process, a portion of the final internal state as a hash of the input.

Abstract: Systems and methods generate reasonably secure hash values at relatively few CPU cycles per byte. An example method includes, for each of a plurality of packets, injecting the packet into an internal state that represents an internal hash sum, mixing the internal state using multiplication, and shuffling the result of the multiplication so that bytes with highest quality are moved to locations that will propagate most widely in a next multiplication operation. Each of the plurality of packets include data from an input to be hashed. In some implementation, a last packet for the input is padded. The method may also include further mixing the internal state using multiplication after processing the plurality of packets and providing, to a requesting process, a portion of the final internal state as a hash of the input.

Abstract: An encoder and/or a computer implemented encoding method includes a texture module configured to determine texture data associated with texture of an image, a noise module configured to determine noise data based on the texture data, a synthesis module configured to generate spatial spectral characteristics of the noise, and combine at least one of the noise data, the texture data, and the spatial spectral characteristics of the noise based on at least one border between adjacent textures, and an encoding module configured to compress the image using an image compression codec.