Abstract: A poke-in terminal includes a main body, a tab terminal socket at a mating end, and a wire socket at a terminating end. The tab terminal socket includes walls forming a tab terminal receptacle configured to receive a tab terminal. Upper and lower beam contacts extend into the tab terminal receptacle to interface with the tab terminal. The wire socket includes walls forming a wire receptacle configured to receive a wire. Poke-in retention beams extend into the wire receptacle to interface with the wire when poked into the wire socket. The wire socket includes a wire release element configured to be actuated to release the poke-in retention beams from the wire.

Abstract: A shelf system mounts to a wall and has a pair of mounting rails mounted to the wall. The mounting rails are elongated, vertically oriented, and spaced-apart from one another. The mounting rails have an array of bores extending laterally into and arrayed vertically along the mounting rail. A shelf is coupled to and between the pair of mounting rails. Lateral apertures in lateral sides of the shelf receive fasteners extending laterally through the lateral apertures and into the array of bores in the mounting rails. The shelf can be formed of a single piece of sheet material bent to form a horizontal platform, lateral sides, and longitudinal tabs.

Abstract: A vertical mount system for wall hangings comprises a wall mounting fastener alignment plate with an array of apertures to align with a vertical array of wall anchors in the wall and secured by a vertical array of double-headed fasteners. The double-headed fasteners extend through the apertures in the alignment plate and into the wall anchors. The alignment plate positions the wall anchors with respect to the wall. A mounting rail is attached to the alignment plate by the double-headed fasteners. A recess in the rear of the mounting rail receives the alignment plate therein. An array of keyholes in the mounting rail receives outer heads of the double-headed fasteners.

Abstract: A modular floating shelf system comprises a modular floating shelf bracket mounting a shelf to a wall and with the modular floating shelf bracket being contained in and hidden by the shelf. The modular floating shelf bracket comprises at least two bracket modules, including a main bracket and an extension bracket, selectively coupleable together. A coupler plate is coupled to a proximal end of an extension mounting plate and overhangs a seam between the extension mounting plate and a main mounting plate. A bore extends through the overhang of the coupler plate and receives a cantilevered main beams of the main bracket. The main and extension brackets can be selectively coupled together to accommodate wider shelves.

Abstract: A poke-in terminal includes a main body, a tab terminal socket at a mating end, and a wire socket at a terminating end. The tab terminal socket includes walls forming a tab terminal receptacle configured to receive a tab terminal. Upper and lower beam contacts extend into the tab terminal receptacle to interface with the tab terminal. The wire socket includes walls forming a wire receptacle configured to receive a wire. Poke-in retention beams extend into the wire receptacle to interface with the wire when poked into the wire socket. The wire socket includes a wire release element configured to be actuated to release the poke-in retention beams from the wire.

Abstract: A modular floating shelf system comprises a modular floating shelf bracket mounting a shelf to a wall and with the modular floating shelf bracket being contained in and hidden by the shelf. The modular floating shelf bracket comprises at least two bracket modules, including a main bracket and an extension bracket, selectively coupleable together. A coupler plate is coupled to a proximal end of an extension mounting plate and overhangs a seam between the extension mounting plate and a main mounting plate. A bore extends through the overhang of the coupler plate and receives a cantilevered main beams of the main bracket. The main and extension brackets can be selectively coupled together to accommodate wider shelves.

Abstract: A print is hung on a wall with a print hanger that has a bracket and a cover plate. The bracket is mounted to a wall and carries the cover plate. The cover plate has magnets that align with lobes of the bracket to clamp an edge of the print and secure the cover plate to the bracket. The cover plate has a pocket to receive and hide the bracket.

Abstract: Systems and methods are disclosed for generating internal state representations of a neural network during processing and using the internal state representations for classification or search. In some embodiments, the internal state representations are generated from the output activation functions of a subset of nodes of the neural network. The internal state representations may be used for classification by training a classification model using internal state representations and corresponding classifications. The internal state representations may be used for search, by producing a search feature from an search input and comparing the search feature with one or more feature representations to find the feature representation with the highest degree of similarity.

Abstract: A poke-in terminal includes a main body, a tab terminal socket at a mating end, and a wire socket at a terminating end. The tab terminal socket includes walls forming a tab terminal receptacle configured to receive a tab terminal. Upper and lower beam contacts extend into the tab terminal receptacle to interface with the tab terminal. The wire socket includes walls forming a wire receptacle configured to receive a wire. Poke-in retention beams extend into the wire receptacle to interface with the wire when poked into the wire socket. The wire socket includes a wire release element configured to be actuated to release the poke-in retention beams from the wire.

Abstract: Systems and methods are disclosed for generating internal state representations of a neural network during processing and using the internal state representations for classification or search. In some embodiments, the internal state representations are generated from the output activation functions of a subset of nodes of the neural network. The internal state representations may be used for classification by training a classification model using internal state representations and corresponding classifications. The internal state representations may be used for search, by producing a search feature from an search input and comparing the search feature with one or more feature representations to find the feature representation with the highest degree of similarity.

Abstract: A poke-in terminal includes a main body, a tab terminal socket at a mating end, and a wire socket at a terminating end. The tab terminal socket includes walls forming a tab terminal receptacle configured to receive a tab terminal. Upper and lower beam contacts extend into the tab terminal receptacle to interface with the tab terminal. The wire socket includes walls forming a wire receptacle configured to receive a wire. Poke-in retention beams extend into the wire receptacle to interface with the wire when poked into the wire socket. The wire socket includes a wire release element configured to be actuated to release the poke-in retention beams from the wire.

Abstract: Systems and methods are disclosed for end-to-end neural networks for speech recognition and classification and additional machine learning techniques that may be used in conjunction or separately. Some embodiments comprise multiple neural networks, directly connected to each other to form an end-to-end neural network. One embodiment comprises a convolutional network, a first fully-connected network, a recurrent network, a second fully-connected network, and an output network. Some embodiments are related to generating speech transcriptions, and some embodiments relate to classifying speech into a number of classifications.

Abstract: Food waste disposer systems and related systems for coupling food waste disposer systems having permanent magnet (PM) motors to electrical power sources, and related coupling and configuration methods, are disclosed herein. In at least one example embodiment, a food waste disposer system includes a PM motor, an AC-to-DC converter having AC input and DC output ports, a housing, a power cord connection structure supported at least indirectly upon the housing, an adapter, and a power cord including a first plug configured to be coupled to the connection structure. The disposer system can be selectively implemented either in a first installation environment including a wall outlet by way of coupling the first plug to the connection structure, or in a second installation environment including a hardwired power cable by coupling additional lead wires of the hardwired power cable to the adapter and further coupling the adapter to the connection structure.

Abstract: Food waste disposer systems and related systems for coupling food waste disposer systems to electrical power sources, and related coupling and configuration methods, are disclosed herein. In at least one example embodiment, a food waste disposer system includes a motor, a switch module, a housing, and an appliance inlet power cord connection structure supported at least indirectly upon the housing, where the connection structure is either connected to or integrally formed with the switch module. Further, the food waste disposer system also includes a power cord including a first end with a first plug and a second end with a second plug, where the first plug is configured to be coupled to the connection structure. Additionally, the food waste disposer includes an adapter or first lead wires that are coupled at least indirectly to the connection structure.

Abstract: Systems and methods are disclosed for generating internal state representations of a neural network during processing and using the internal state representations for classification or search. In some embodiments, the internal state representations are generated from the output activation functions of a subset of nodes of the neural network. The internal state representations may be used for classification by training a classification model using internal state representations and corresponding classifications. The internal state representations may be used for search, by producing a search feature from an search input and comparing the search feature with one or more feature representations to find the feature representation with the highest degree of similarity.

Abstract: Systems and methods are disclosed for generating internal state representations of a neural network during processing and using the internal state representations for classification or search. In some embodiments, the internal state representations are generated from the output activation functions of a subset of nodes of the neural network. The internal state representations may be used for classification by training a classification model using internal state representations and corresponding classifications. The internal state representations may be used for search, by producing a search feature from an search input and comparing the search feature with one or more feature representations to find the feature representation with the highest degree of similarity.

Abstract: Systems and methods are disclosed for end-to-end neural networks for speech recognition and classification and additional machine learning techniques that may be used in conjunction or separately. Some embodiments comprise multiple neural networks, directly connected to each other to form an end-to-end neural network. One embodiment comprises a convolutional network, a first fully-connected network, a recurrent network, a second fully-connected network, and an output network. Some embodiments are related to generating speech transcriptions, and some embodiments relate to classifying speech into a number of classifications.

Abstract: Systems and methods are disclosed for end-to-end neural networks for speech recognition and classification and additional machine learning techniques that may be used in conjunction or separately. Some embodiments comprise multiple neural networks, directly connected to each other to form an end-to-end neural network. One embodiment comprises a convolutional network, a first fully-connected network, a recurrent network, a second fully-connected network, and an output network. Some embodiments are related to generating speech transcriptions, and some embodiments relate to classifying speech into a number of classifications.

Abstract: Systems and methods are disclosed for generating internal state representations of a neural network during processing and using the internal state representations for classification or search. In some embodiments, the internal state representations are generated from the output activation functions of a subset of nodes of the neural network. The internal state representations may be used for classification by training a classification model using internal state representations and corresponding classifications. The internal state representations may be used for search, by producing a search feature from an search input and comparing the search feature with one or more feature representations to find the feature representation with the highest degree of similarity.

Abstract: Systems and methods are disclosed for end-to-end neural networks for speech recognition and classification and additional machine learning techniques that may be used in conjunction or separately. Some embodiments comprise multiple neural networks, directly connected to each other to form an end-to-end neural network. One embodiment comprises a convolutional network, a first fully-connected network, a recurrent network, a second fully-connected network, and an output network. Some embodiments are related to generating speech transcriptions, and some embodiments relate to classifying speech into a number of classifications.