Abstract: A method of forming an insulated composite fabric including attaching an insulating-filler fabric layer to an outer fabric layer, where the insulating-filler fabric layer is a textile fabric with at least one raised surface on the fabric. One side of the insulating-filler layer has a first surface with relatively high pile, including regions of no pile or relatively low pile interspersed among regions of relatively high pile and the other side of the insulating-filler layer has a second surface with relatively high pile, including regions of no pile or relatively low pile interspersed among regions of relatively high pile. The method further contains coupling at least one of a fleece and velour that is respectively lower than the relatively high pile of the first surface or the relatively high pile of the second surface to at least one of the first surface and the second surface.

Abstract: A computer-implemented intent classification method and system are provided. The method comprises: providing a plurality of trained model sets, each based on a different model and comprising a plurality of trained models, with each trained model in each trained model set obtained by training a same model with a different initialization condition over a pre-determined dataset; feeding an input text into each trained model to obtain a plurality of prediction results for indicating a likely intent class of the input text; ensembling prediction results corresponding to a same trained model set for each trained model set to obtain a plurality of first-layer ensembles; ensembling the plurality of first-layer ensembles to obtain at least one second-layer ensemble, each obtained by ensembling two or more first-layer ensembles; and determining the intent class of the input text based on the at least one second-layer ensemble.

Abstract: A voice activity detection method includes: training one or more computerized neural networks having a denoising autoencoder and a classifier, wherein the training is performed utilizing one or more models including Mel-frequency cepstral coefficients (MFCC) features, ? features, ?? features, and Pitch features, each model being recorded at one or more differing associated predetermined signal to noise ratios; recording a raw audio waveform and transmitting the raw audio waveform to the computerized neural network; denoising the raw audio wave utilizing the denoising autoencoder; and determining whether the raw audio waveform contains human speech; extracting any human speech from the raw audio waveform.

Abstract: A computer-implemented intent classification method and system are provided. The method comprises: providing a plurality of trained model sets, each based on a different model and comprising a plurality of trained models, with each trained model in each trained model set obtained by training a same model with a different initialization condition over a pre-determined dataset; feeding an input text into each trained model to obtain a plurality of prediction results for indicating a likely intent class of the input text; ensembling prediction results corresponding to a same trained model set for each trained model set to obtain a plurality of first-layer ensembles; ensembling the plurality of first-layer ensembles to obtain at least one second-layer ensemble, each obtained by ensembling two or more first-layer ensembles; and determining the intent class of the input text based on the at least one second-layer ensemble.

Abstract: Methods and apparatuses are described for intelligent automation of computer software test scripts and code requirements. A server generates a baseline test script for a plurality of browser-based applications with a sequence of action steps having one or more empty user interface component values. The server performs a first execution of the baseline test script against each of the browser-based applications to extract one or more expected values associated with user interface components that correspond to the empty user interface component values, and stores the expected values. The server performs a second execution of the baseline test script against the browser-based applications to extract one or more actual values associated with user interface components that correspond to the empty user interface component values. The server compares the actual values against the expected values, and if the values do not match, the server generates a message indicating a test failure.

Abstract: A pet toy is provided that is especially adapted for tug-of-war play between a pet owner and animal. The pet toy includes a pair of halves or parts with a handle on each part. A hinge interconnects the pair of parts. The hinge is configured in a shape and made of a material such that a desired amount of torque can be transmitted from one end of the pet toy to another when the pet owner and animal engage in a tug-of-war action. The handle is configured and made of material that minimizes damage to the teeth and gums of the animal's mouth, and also reduces discomfort to the hand of the pet owner. The handles are arranged substantially perpendicular to a direction of longitudinal pull which occurs in the tug-of-war action.

Abstract: A treat-dispensing toy for pets. The embodiment may dispense a substance, such as a treat, when a pet interacts with the embodiment. The embodiment may include a body; an interior chamber defined within the body; a first planar surface formed within the interior chamber; an opening formed in the body and communicating with the interior chamber; wherein the interior chamber is adapted to hold the substance; and the opening is adapted to pass the substance to the interior chamber. Further, the embodiment may include a passage extending from the opening to the interior chamber, the passage at least partially defined by the first planar surface. The embodiment may also include a second planar surface formed within the interior chamber; and a second opening formed in the body and communicating with the interior chamber.

Abstract: A pet toy having a substructure and a resilient material bonded to at least a portion of the substructure, wherein the resilient material defines a bite zone sized to accept at least a portion of an animal jaw. Further, the substructure may include an interior circumference, an exterior circumference, and at least one rib joining the interior circumference to the exterior circumference. The interior circumference, exterior circumference, and at least one rib at least may partially define the bite zone, and the bite zone may be at least partially covered by the resilient material. The embodiment may be disc-shaped.

Abstract: The time management system for mobile employees is a telecommunications system allowing a company client to manage their off-site employees efficiently and effectively. Mobile employees, in communication with the system Web server via a cell phone or PDA, enter time transaction data using the mobile computing device's Web browser. Based upon the employee's predefined profile, custom Web pages are presented requesting employee specific time transaction information. Once received, time transaction data may be forwarded to the company client's computer directly or downloaded to client computers upon request or based upon a predetermined schedule. Having at least one Web server and application/database server, the system maintains employee data, client data, project data, and transaction data. The system protects against unauthorized access of data by having defined administrative user privileges.

Abstract: A powdered lubricant applicator is provided which utilizes a venturi nozzle to draw powder from a fluidized hopper. The amount of powder delivered from the hopper is controlled by the air flow rate through the nozzle and by the length of time that the air flows through the nozzle. As air passes through the nozzle, powdered lubricant is drawn from the fluidized hopper and passed with the air along a delivery tube. As the fluidized powdered lubricant passes along the delivery tube, additional "atomizing" air is introduced into the delivery tube to assist powder delivery to the destination mechanism to be lubricated. In addition, fluidizing air is added to the delivery tube. The air used to assist the powder delivery (i.e., the atomizing air) and the air used to fluidize the powder (i.e., the fluidizing air) impact the powder delivery. Three pressure regulators are used to independently control the flow air, the atomizing air and the fluidizing air.