Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for compressing and decompressing data. In one aspect, a method comprises: processing data using an encoder neural network to generate a latent representation of the data; processing the latent representation of the data using a hyper-encoder neural network to generate a latent representation of an entropy model; generating an entropy encoded representation of the latent representation of the entropy model; generating an entropy encoded representation of the latent representation of the data using the latent representation of the entropy model; and determining a compressed representation of the data from the entropy encoded representations of: (i) the latent representation of the data and (ii) the latent representation of the entropy model used to entropy encode the latent representation of the data.

Abstract: A tap assembly includes a flow control mechanism, a spindle which rotates about its principal axis to operate the flow control mechanism, and a handle connectable to the spindle. The location on the handle where the spindle connects coincides with the spindle's principal axis but does not with the centroid of the handle's planform shape. When the handle is turned by a user, the handle rotates about the spindle's principal axis. The handle also moves or translates relative the spindle's principal axis. When the handle is in an initial “fully off” position, there is an area that is obscured from a user's view by the handle, but when the handle is initially turned from the initial position towards a final “fully on” position, the area begins to be revealed, and with further rotation of the handle towards the final position, more of, or different parts of, the area become revealed.

Abstract: A scalable enterprise platform for automated functional and integration regression testing is provided. Embodiments of the disclosed system facilitate the testing of any number of different software systems in development, even where the systems have unique dataset formats. Embodiments of the present invention provide a common method to generate logging and results reports across the platform, thereby providing simpler results analysis. Embodiments may also standardize the query set and facilitate the capability to analyze large results sets. Furthermore, embodiments of the disclosed system may combine the original data to the validated data to allow testers to analyze the testing results. In addition, embodiments of the present invention supports secured separation of testing domains. In at least one embodiment, the system includes a centralized user interface system that provides users with different domains to securely access one or more testing domains.

Abstract: Methods, and systems, including computer programs encoded on computer storage media for compressing data items with variable compression rate. A system includes an encoder sub-network configured to receive a system input image and to generate an encoded representation of the system input image, the encoder sub-network including a first stack of neural network layers including one or more LSTM neural network layers and one or more non-LSTM neural network layers, the first stack configured to, at each of a plurality of time steps, receive an input image for the time step that is derived from the system input image and generate a corresponding first stack output, and a binarizing neural network layer configured to receive a first stack output as input and generate a corresponding binarized output.

Abstract: A reading group that is associated with an ebook is defined, the reading group including a master device and one or more follower devices. A synchronization command is received from the master device to synchronize the reading location of the one or more follower devices to a read section, the read section specifying a portion of the ebook. The synchronization command is provided to the one or more follower devices, and in response to receiving the synchronization command the one or more follower devices are configured to present the ebook in accordance with the read section.

Abstract: An image sensor has a plurality of pixels arranged in an array, a selection means for selecting individual pixels in the array, and a shutter means for transmitting a shutter signal to the pixels. The pixels are constructed and arranged to sense incident light only when both selected and in receipt of a shutter signal. The pixels in said array can thus be triggered individually to capture light at different times.

Abstract: A reading group that is associated with an ebook is defined, the reading group including a master device and one or more follower devices. A synchronization command is received from the master device to synchronize the reading location of the one or more follower devices to a read section, the read section specifying a portion of the ebook. The synchronization command is provided to the one or more follower devices, and in response to receiving the synchronization command the one or more follower devices are configured to present the ebook in accordance with the read section.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for classifying voice inputs. The methods, systems, and apparatus include actions of providing an answer to a first voice input from a user and receiving visual or audio data corresponding to a second voice input. Further actions include classifying the second voice input as a follow on request to the first voice input or as deliberation on the answer, based on the visual data or the audio data. Additionally, the actions include determining whether to provide a response to the second voice input based on the classification of the second voice input.

Abstract: A method and apparatus for asynchronous analysis of URLs in messages in a live message processing environment whereby an identified URL included in a message identified for further analysis is extracted and prioritized based on one or more prioritization parameters. The URL is then sent from a live message processing stream to a URL analysis process and the live message processing stream is stalled, with respect to that message, until the results of the URL analysis process are received, or until a defined “timeout” period passes. The results, or “verdict”, of the URL analysis process are then sent back to the live message processing stream in an asynchronous manner based on the prioritization of the URL. If the results of the URL analysis process are not received by the live message processing stream before the defined “timeout” period passes, then the “best” action is taken based on the best available data/verdict.

Abstract: JavaScript on webpages linked to by URLs in messages is identified and the JavaScript is extracted. The JavaScript is then subjected to a JavaScript execution and analysis process whereby the JavaScript is executed in the context of a simulated web browser. The behavior of the JavaScript is then analyzed to identify one or more of: any URLs to be redirected to; any further executable JavaScript; and any content dynamically written to the webpage. The results are then either recursed into or are recorded and used to aid in the identification of spam messages.

Abstract: A method and apparatus for automatically obtaining web page content in the presence of redirects whereby an incoming message is received and analyzed to determine if the message contains any URLs. Any detected URLs are then extracted and activated to analyze the contents of the web page linked to by the URL. The HTTP response headers and content sent from a web page server in response to the browser HTTP requests to activate the URL link are analyzed to determine if the response includes a redirect to a new, or destination, URL, and associated web page, i.e., to determine if the detected URLs result in redirects. If the HTTP response indicates a redirect, a URL redirect analysis process is initiated that includes a set of redirect processing procedures that are selectively applied depending on the type of redirect encountered, and each redirect is automatically followed. For chains of redirects, the process is recursive, i.e.

Abstract: An image sensor has a plurality of pixels arranged in an array, a selection means for selecting individual pixels in the array, and a shutter means for transmitting a shutter signal to the pixels. The pixels are constructed and arranged to sense incident light only when both selected and in receipt of a shutter signal. The pixels in said array can thus be triggered individually to capture light at different times.

Abstract: A computer-implemented method for creating text signatures for identifying spam messages may include: 1) identifying a spam message, 2) tokenizing the text of the spam message, 3) identifying at least one incorrectly punctuated token within the tokenized text of the spam message, 4) identifying, within the tokenized text of the spam message, at least one token that is adjacent to the incorrectly punctuated token, and then 5) creating a candidate signature for identifying future instances and/or variations of the spam message that includes both the incorrectly punctuated token and the token that is adjacent to the incorrectly punctuated token. Corresponding systems and configured computer-readable media are also disclosed.

Abstract: A method and apparatus for identifying scam e-mails using phone number analysis whereby a set of potential scam e-mail identification rules is created and data representing the potential scam e-mail identification rules is stored in a data storage device. Incoming e-mails are then scanned under the direction of one or more processors by an e-mail security system to detect phone numbers present in the incoming e-mails using one or more phone number detection parameters. Any potential phone numbers detected using the one or more phone number detection parameters are then normalized. In one embodiment, the normalized potential phone numbers are then analyzed using the potential scam e-mail identification rules to identify potential scam phone numbers. The e-mails including the identified potential scam phone numbers then have their status transformed to that of potential scam e-mails.