摘要:
A method for manufacturing three-dimensional surface figures on a soft magnetic material. An insulating layer is first laid on a soft magnetic material layer and a foam intermediate layer is then laid on the insulating layer. A face layer previously printed with figures is finally laid on the intermediate layer. Thereafter the face layer is embossed and molded with concave/convex three-dimensional figures corresponding to the present embossing stripes of a molding device by a high frequency device so as to provide an improvement over conventional three-dimensional figures made by injection molding which cannot have figures printed by halftone block printing or color separation so as to eliminate the shortcomings existing in planar printed magnetic products which are only printed with planar figures.
摘要:
Weights can be pruned during DNN training to increase sparsity in the weights and reduce the amount of computation required for performing the deep learning operations in DNNs. A DNN layer may have one or more weight tensors corresponding to one or more output channels of the layer. A weight tensor has weights, the values of which are determined by training the DNN. A weight tensor may have a dimension corresponding to the input channels of the layer. The weight tensor may be partitioned into subtensors, each of which has a subset of the input channels. The subtensor may have the same number of input channels. One or more subtensors may be selected, e.g., based on the weights in the one or more subtensors. The weights in a selected subtensor are pruned, e.g., changed to zeros. The weights in an unselected subtensor may be modified by further training the DNN.
摘要:
This disclosure is directed to a digital signal processing (DSP) block that includes multiple weight registers configurable to receive and store a first plurality of values, and multiple multipliers that are each configurable to receive a respective value of the first plurality of values. The DSP block further includes one or more inputs configurable to receive a second plurality of values, and a multiplexer network configurable to receive the second plurality of values and route each respective value of the second plurality of values to a multiplier of the multipliers. The multipliers are configurable to simultaneously multiply each value of the first plurality of values by a respective value of the second plurality of values to generate a plurality of products. Additionally, the DSP block includes adder circuitry configurable to generate a first sum and a second sum based on the plurality of products.
摘要:
A device for removing a composition from the skin, for example a depilatory composition, has a handle (1), which carries a non-shaving head (2) which is moved over the skin to effect removal of the composition. There is a zone of flexure (5) which extends laterally across the head separating it into a front portion (4) and a rear portion (3). There may be additional zones of flexure within the device.
摘要:
Embodiments of tie-down apparatuses and associated methods are disclosed herein. An exemplary tie-down/sling apparatus can include a first flat component having a first surface and a second flat component having a first side and a second side opposite the first side. The second flat component can be positioned on the first flat component by fixedly attaching the first side to the first surface. The second side of the second flat component can include a reflective surface configured to visually present a set of information to an operator or others.
摘要:
A device for removing a composition from the skin (e.g. a depilatory composition) has a handle (44) and a non-shaving head (46), having an under-surface from which one or more fins (48) projects transversely. The device is held by the handle (44) and moved over the skin so that the fin(s) and the head can remove the composition.
摘要:
A computer server system provides an online community for customers of a business entity. A first level feedback module provides visual feedback to each user in a user interface for the online community, which quantifies each user's interaction with the community. A notification module sends an electronic message to a user when the user's interaction is below a predefined threshold rate. A second level feedback module calculates second level metrics that quantify other users' evaluations of postings to the online community by a first user. Visual feedback of the second level metrics is provided. A third level feedback module awards a designated achievement badge to users whose interaction metrics exceed predefined threshold values, and displays the achievement badge in the user interface. The system promotes a high level of user participation in the online community and thereby contributes to a low cost of technical support provided by the business entity.
摘要:
A method of processing a signal within a receiver can include generating, using the receiver, a candidate list including at least one entry for a signal. Each entry can include a candidate symbol vector and an L1-Norm error metric. The method can include, for each entry, generating an L1-Norm transformation from the L1-Norm error metric, wherein the L1-Norm transformation depends upon a function of a number of receiving antennas of the receiver. The signal can be decoded using the candidate list including the L1-Norm transformations.
摘要:
A method of load distribution for a cluster of two or more nodes. The method comprises receiving an initial request packet on a network device having a virtual IP address; forwarding the request packet from the network device to a cluster of at least two nodes, wherein each of the at least two nodes has an internal dispatcher module and an unique and non-conflicting virtual IP address; establishing one of the at least two nodes as a priority dispatcher or dispatcher endpoint, wherein if any one node fails, the virtual IP address of the one node which is no longer active falls back to another node within the cluster based on cluster priorities; dispatching the request packet to one of the nodes associated with the cluster; and forwarding the request from one of the nodes to a switching device.
摘要:
A method of application level content routing using a (Dynamic VPN) dual-proxy mechanism that provides a client access to resources on a remote network without exposing each other's actual IP addresses and network topologies. The method includes providing a client a list of available resources on a remote network; initiating a request by the client for at least one resource from the list of available resources hosted on the remote network as though the at least one resource is local to the client; routing the request to the at least remote resource through a secure connection between the client and the remote network; responding to the request by the at least remote resource as though the request is initiated locally on the remote network; and routing the response from the remote network back to the client through the secure connection.