Abstract: A method for automatically creating an equipment inventory from a plurality of images of a site. The method includes receiving, by a processor, the plurality of images of the site, labeling, by the processor, equipment present in each image of the plurality of images of the site, creating, by the processor, an overlap curve for neighboring images of the plurality of images of the site, determining, by the processor, a useful image of each side of the site based on the overlap curve, the useful image including a subset of the plurality of images, creating, by the processor, an overlay of equipment in the useful image of each side of the site, counting, by the processor, equipment present in the overlay of each side of the site, and generating, by the processor, the equipment inventory by adding a number of equipment counted in the overlay of each side of the site.

Abstract: An apparatus includes at least one processor; and at least one non-transitory memory including computer program code; wherein the at least one memory and the computer program code are configured to, with the at least one processor, cause the apparatus at least to: estimate an importance of parameters of a neural network based on a graph diffusion process over at least one layer of the neural network; determine the parameters of the neural network that are suitable for pruning or sparsification; remove neurons of the neural network to prune or sparsify the neural network; and provide at least one syntax element for signaling the pruned or sparsified neural network over a communication channel, wherein the at least one syntax element comprises at least one neural network representation syntax element.

Abstract: Apparatus and method of evaluating electronic documents. In an embodiment, the method comprises obtaining a dataset comprising labeled text samples associated with subject matter categories and having a respective label comprises a risk level. The method comprises applying a machine learning (ML) model to the labeled text samples to vectorize the labeled text samples, and determining a representative vector for groups of text sample vectors associated with a same risk level in each subject matter category. The method comprises displaying an electronic document, extracting a text segment, applying the ML model to the text segment to vectorize the text segment, mapping the text segment vector to a subject matter category, determining the risk level associated with the text segment based on a relation between the text segment vector and representative vectors associated with the subject matter category, and annotating the text segment with an annotation based on the risk level.

Abstract: A method for automatically creating an equipment inventory from a plurality of images of a site. The method includes receiving, by a processor, the plurality of images of the site, labeling, by the processor, equipment present in each image of the plurality of images of the site, creating, by the processor, an overlap curve for neighboring images of the plurality of images of the site, determining, by the processor, a useful image of each side of the site based on the overlap curve, the useful image including a subset of the plurality of images, creating, by the processor, an overlay of equipment in the useful image of each side of the site, counting, by the processor, equipment present in the overlay of each side of the site, and generating, by the processor, the equipment inventory by adding a number of equipment counted in the overlay of each side of the site.

Abstract: Systems, methods, and software of processing Deep Neural Networks (DNN) comprising a plurality of nodes with connections between the nodes. In one embodiment, a system initializes the DNN by assigning initial values to weights associated with the connections, and identifies an initial subset of the connections randomly selected from the DNN. The initial subset after training with a training dataset has a training accuracy below an accuracy threshold. The system performs a growth process by adding additional connections from the DNN to the initial subset to generate a qualified subset of the connections. The qualified subset of the connections after training with the training dataset has a training accuracy that reaches the accuracy threshold.

Abstract: An apparatus includes at least one processor; and at least one non-transitory memory including computer program code; wherein the at least one memory and the computer program code are configured to, with the at least one processor, cause the apparatus at least to: estimate an importance of parameters of a neural network based on a graph diffusion process over at least one layer of the neural network; determine the parameters of the neural network that are suitable for pruning or sparsification; remove neurons of the neural network to prune or sparsify the neural network; and provide at least one syntax element for signaling the pruned or sparsified neural network over a communication channel, wherein the at least one syntax element comprises at least one neural network representation syntax element.

Abstract: The method includes compiling data into mutual information columns, determining mutual information for each pairing of the mutual information columns and creating a matrix using the mutual information, the matrix including a first set of data columns, wherein each of the first set of data columns represents at least one feature of the data. The method further includes computing eigenvalues and eigenvectors of the matrix, ordering the eigenvalues using an absolute value of the eigenvalues, iteratively selecting at least one second set of data columns by successively removing data columns from the first set of data columns based on the ordered eigenvalues, and controlling an operation of an electronic device based on the at least one second set of data.

Abstract: The method includes compiling data into mutual information columns, determining mutual information for each pairing of the mutual information columns and creating a matrix using the mutual information, the matrix including a first set of data columns, wherein each of the first set of data columns represents at least one feature of the data. The method further includes computing eigenvalues and eigenvectors of the matrix, ordering the eigenvalues using an absolute value of the eigenvalues, iteratively selecting at least one second set of data columns by successively removing data columns from the first set of data columns based on the ordered eigenvalues, and controlling an operation of an electronic device based on the at least one second set of data.

Abstract: Systems and methods for exploring a data collection. An exemplary system includes a GUI that displays a window to the user, and displays data element icons within the window representing an initial set of data elements from the data collection. Visual properties of the data element icons indicate the relevance of the data elements in the initial set to the data collection. The GUI receives input from the user selecting a data element icon within the window. Selection of the data element icon assigns a corresponding data element in the initial set as an initial anchor set. The GUI displays another window to the user, displays an anchor icon for the initial anchor set, and displays the data element icons representing a relevant set of data elements. Visual properties of the data element icons indicate the relevance of the data elements in the relevant set to the initial anchor set.

Abstract: A memory is configured to store a dataset and a processor is configured to map the dataset to a plurality of objects. The objects are represented by corresponding values of a plurality of non-negative elements. The processor is also configured to construct a bipartite graph including a plurality of first nodes associated with the plurality of objects and a plurality of second nodes associated with the plurality of non-negative elements. The first nodes are linked to the second nodes by edges having weights equal to values of the non-negative elements that represent the corresponding first node. The processor is further configured to determine similarity values that indicate degrees of similarity between the plurality of objects based on a diffusion of a fluid mass through the bipartite graph according to the weights of the edges.

Abstract: A method, apparatus, and machine readable storage medium is disclosed for balancing loads among a plurality of virtual machines (VMs) from a central dispatcher, wherein the dispatcher receives data packets and maps the data packets to VMs selected from the plurality of VMs, using a weighted hash function, having an associated weighting for each VM and forwarding each packet to a VM accordingly, wherein a load balancer decrements a weighting for a VM, responsive to an indication of the load on the VM exceeding a first load threshold. Weightings can correspond to a number of bins associated with each VM. Weightings are adjusted in response to receiving invite and disinvite messages from the VMs, representing their respective loads.

Abstract: Systems and methods are provided for ranking electronic information based on determined similarities. In one aspect a set of unique features are determined from a collection of electronic objects. A graph is constructed in which electronic object are represented as object nodes and determined features are represented as feature nodes. The object nodes are interconnected by a weighted edge to at least one feature node. Scores for the object nodes and the feature nodes are computed using a determined set of anchor nodes and a determined weighted adjacency matrix. The object nodes and the feature nodes of the graph are ranked and displayed based on the computed scores. In one aspect, the scores and the ranks for the object nodes and the feature nodes are dynamically updated and displayed based on user preferences.

Abstract: A method, apparatus, and machine readable storage medium is disclosed for balancing loads among a plurality of virtual machines (VMs) from a central dispatcher, wherein the dispatcher receives data packets and maps the data packets to VMs selected from the plurality of VMs, using a weighted hash function, having an associated weighting for each VM and forwarding each packet to a VM accordingly, wherein a load balancer decrements a weighting for a VM, responsive to an indication of the load on the VM exceeding a first load threshold. Weightings can correspond to a number of bins associated with each VM. Weightings are adjusted in response to receiving invite and disinvite messages from the VMs, representing their respective loads.

Abstract: Systems and methods for exploring a data collection. An exemplary system includes a GUI that displays a window to the user, and displays data element icons within the window representing an initial set of data elements from the data collection. Visual properties of the data element icons indicate the relevance of the data elements in the initial set to the data collection. The GUI receives input from the user selecting a data element icon within the window. Selection of the data element icon assigns a corresponding data element in the initial set as an initial anchor set. The GUI displays another window to the user, displays an anchor icon for the initial anchor set, and displays the data element icons representing a relevant set of data elements. Visual properties of the data element icons indicate the relevance of the data elements in the relevant set to the initial anchor set.

Abstract: System comprising a processing unit configured to compute transmit signals and an array of speakers in wired or wireless communication with the processing unit, each of the speakers of the array at different locations in a common setting. The speakers, upon receiving the transmit signals, simultaneously transmit both a first audio and a different second audio stream into the setting from the speakers. The first stream from the speakers aggregate in the vicinity of a first location in the setting to form an aggregated first stream that is audible to human hearing. The second stream transmitted from the speakers do not aggregate at the first location. The second stream from the speakers aggregate in the vicinity of a different second location in the setting to form an aggregated second stream that is audible to human hearing. The first stream transmitted from the speakers do not aggregate at the second location.

Abstract: Various exemplary embodiments relate to a chaining of sequential functions associated with a service or application are considered. This approach relies on a centralized load balancer for reducing the load of inter-rack traffic in a data center. The centralized load balancer may include a memory configured to store a service data flow table; and a processor configured to: receive at the centralized load balancer, a path inquiry for a service data flow; determine which virtual machine to assign the service data flow, wherein at least two functions of a chain of functions required in the service data flow are to be performed on the same rack; and assign the service data flow to the determined virtual machine.

Abstract: In one example embodiment, a system for localizing an audio stream includes a processor. The processor is configured to determine channel state information of an acoustic channel between a plurality of speakers and at least one device of a plurality of devices, the at least one device requesting the audio stream from among available audio streams. The processor is further configured to determine transmit signals for transmitting audio signals representing the available audio streams to the plurality of devices, the determined transmit signals being based on at least the determined channel state information such that the requested audio stream is more audible to a user associated with the at least one device compared to other users associated with other ones of the plurality of devices.

Abstract: A capability is provided for reducing response variance of virtual processors. A controller receives a processing request. The controller may propagate the processing request toward multiple virtual processors hosted on multiple hardware devices contemporaneously. The controller may propagate the processing request toward a first virtual processor hosted on a first hardware device and propagate the processing request toward a second virtual processor hosted on a second hardware device based on a determination that a timeout period expires before a processing response is received from the first virtual processor. The timeout period may be determined based on one or more response time statistics of the virtual processor and one or more response time statistics of a physical processor.

Abstract: Delay-sensitive burst scheduling in a network is carried out through communications between source and destination nodes. A distributed scheduler at each destination node selects timeslots for burst arrivals from requesting source nodes, irrespective of timeslot selections made by distributed schedulers at other nodes.

Abstract: Systems and methods are provided for constructing a collection of one or more tree sub-graph representations of a graph including multiple interconnected nodes, where the one or more tree sub-graph representations of the graph are used to estimate the shortest distance between any two nodes of a graph. One of the features of the systems and methods disclosed herein is a methodology for the selection or designation of root nodes for constructing the collection of the one or more tree sub-graph representations. Another feature of the present disclosure is a methodology of expanding the parent nodes in a given level of the tree sub-graph representations into one or more child nodes in a successive level of the tree-graph representations based on a descending order of degree of the parent nodes.