Abstract: The present disclosure relates to the technical field of growth environment control for plants, and provides a light regulation method, system, and apparatus for a growth environment of leafy vegetables. The method includes: obtaining a growth environment image and growth environment light parameters of leafy vegetables; inputting the growth environment image into an image segmentation model to obtain an initial pixel segmentation map; determining a leaf area of the leafy vegetables according to the initial pixel segmentation map by using a Class Activation Mapping (CAM)-K algorithm; inputting the leaf area and the growth environment light parameters of the leafy vegetables into a leafy vegetable growth index prediction model to obtain a growth index of the leafy vegetables; adjusting the growth environment light parameters according to the growth index of the leafy vegetables. This disclosure achieves automated and intelligent control of light conditions in the growth environment of leafy vegetables.

Abstract: Predicting power usage of a chip may include receiving placement data describing a placement, within the chip, of a plurality of logical components of the chip; providing the placement data as an input to a neural network; and determining, by the neural network, based on the placement data, a predicted power usage of the chip.

Abstract: The present disclosure relates to the technical field of aquafarm monitoring, and in particular to a method, system and apparatus for intelligently monitoring an aquafarm with multi-dimensional panoramic perception. The method includes: analyzing preprocessed multi-dimensional panoramic perception information with time series analysis to obtain a linear fitting model; constructing a water quality forecasting-alarming model according to the linear fitting model; inputting fused data to a convolutional neural network (CNN) for training to obtain an electrical device fault diagnosis model; and monitoring an aquafarm with the water quality forecasting-alarming model and the electrical device fault diagnosis model.

Abstract: Predicting power usage of a chip may include receiving placement data describing a placement, within the chip, of a plurality of logical components of the chip; providing the placement data as an input to a neural network; and determining, by the neural network, based on the placement data, a predicted power usage of the chip.

Abstract: Predicting power usage of a chip may include receiving placement data describing a placement, within the chip, of a plurality of logical components of the chip; providing the placement data as an input to a neural network; and determining, by the neural network, based on the placement data, a predicted power usage of the chip.

Abstract: A method a system include obtaining a master list of layer traits including wire codes, each of the wire codes indicating a width of a corresponding wire, and including a maximum reach length of the corresponding wire and a time of flight (TOF) through the corresponding wire. The method also includes processing the master list of the layer traits to obtain a final list of the layer traits, the final list of the layer traits having fewer entries than the master list of the layer traits and being in a ranked order. A metric is calculated for each adjacent pair of the layer traits in the final list of layer traits. The final list of the layer traits and the corresponding metric is used to assign the corresponding wires to different interconnects among components of an integrated circuit.

Abstract: A method a system include obtaining a master list of layer traits including wire codes, each of the wire codes indicating a width of a corresponding wire, and including a maximum reach length of the corresponding wire and a time of flight (TOF) through the corresponding wire. The method also includes processing the master list of the layer traits to obtain a final list of the layer traits, the final list of the layer traits having fewer entries than the master list of the layer traits and being in a ranked order. A metric is calculated for each adjacent pair of the layer traits in the final list of layer traits. The final list of the layer traits and the corresponding metric is used to assign the corresponding wires to different interconnects among components of an integrated circuit.

Abstract: An integrated circuit physical design tool includes an interconnection congestion predictor that generates a congestion table for an integrated circuit design. Interconnection congestion hotspots are identified based on the congestion table. The proximity of interconnection congestion hotspots to each other is determined, and penalization values are computed based on the proximity of interconnection congestion hotspots to each other. The congestion table is then updated to reflect the penalization values due to proximity of interconnection congestion hotspots. Routability of the interconnection congestion hotspots is then predicted based on the updated congestion table. The updated congestion table may also be used by multiple physical design optimization tools, including placement, global routing, and detail routing.

Abstract: Methods and systems for improving the performance of a computer performing an electronic design. One or more nets of a netlist are sorted based on an amount of slack and a net of the one or more nets that is unprocessed and that has a least amount of slack is selected as a current target net. A layer of a higher bucket that is unprocessed for the currently selected target net is selected, the higher bucket being higher than a bucket of the current target net. A determination of whether capacity is available to route the current target net on the selected layer of the higher bucket is made and the current target net is routed on the selected layer of the higher bucket in response to capacity being available. One or more nets that are competing for resources with the current target net on the selected layer of the higher bucket are identified as candidate nets in response to capacity not being available.

Abstract: A wireless charger can determine a power requirement associated with operating at least a first device at a scheduled time. The wireless charger can determine a power emission level for the wireless charger based, at least in part, on the power requirement, such that the power emission level will provide sufficient wireless energy to satisfy the power requirement. The wireless charger can transmit wireless energy at the power emission level to cause the device to operate.

Abstract: Embodiments disclose systems, methods, and computer program products to perform an operation for proactively sending content to mobile devices. The operation generally includes receiving, from an application executing on a mobile device, a request to access content, and identifying, from a plurality of content items, at least a first content item responsive to the request. The operation also includes determining at least a second one of the content items based at least in part on the request, a context of the mobile device and feedback regarding at least one of the content items sent to the mobile device. The operation further includes generating a response to the request that includes at least the first content item and the second content item, and sending the response to the application.

Abstract: A method of performing physical design of an integrated circuit includes subdividing each metal layer of a plurality of metal layers of the integrated circuit into a plurality of g-cells. Each metal layer has either horizontal or vertical tracks, the g-cells of the metal layers with horizontal tracks have vertical edges between adjacent ones of the g-cells, and the g-cells of the metal layers with vertical tracks have horizontal edges between adjacent ones of the g-cells. The method includes determining congestion for each metal layer as congestion values associated with the horizontal edges or the vertical edges of the metal layer, identifying hotspots for each metal layer based on the congestion values of the metal layer, determining a penalty associated with the hotspots of each metal layer, determining a congestion metric for each metal layer based on the penalty, and performing routing of the wires based on the congestion metric.

Abstract: An integrated circuit physical design tool includes an interconnection congestion predictor that generates a congestion table for an integrated circuit design. Interconnection congestion hotspots are identified based on the congestion table. The proximity of interconnection congestion hotspots to each other is determined, and penalization values are computed based on the proximity of interconnection congestion hotspots to each other. The congestion table is then updated to reflect the penalization values due to proximity of interconnection congestion hotspots. Routability of the interconnection congestion hotspots is then predicted based on the updated congestion table. The updated congestion table may also be used by multiple physical design optimization tools, including placement, global routing, and detail routing.

Abstract: Predicting power usage of a chip may include receiving placement data describing a placement, within the chip, of a plurality of logical components of the chip; providing the placement data as an input to a neural network; and determining, by the neural network, based on the placement data, a predicted power usage of the chip.

Abstract: A method of performing physical design of an integrated circuit includes subdividing each metal layer of a plurality of metal layers of the integrated circuit into a plurality of g-cells. Each metal layer has either horizontal or vertical tracks, the g-cells of the metal layers with horizontal tracks have vertical edges between adjacent ones of the g-cells, and the g-cells of the metal layers with vertical tracks have horizontal edges between adjacent ones of the g-cells. The method includes determining congestion for each metal layer as congestion values associated with the horizontal edges or the vertical edges of the metal layer, identifying hotspots for each metal layer based on the congestion values of the metal layer, determining a penalty associated with the hotspots of each metal layer, determining a congestion metric for each metal layer based on the penalty, and performing routing of the wires based on the congestion metric.

Abstract: Systems, methods, and computer program products to perform an operation comprising receiving, by a fog computing platform, a request from a wireless device to perform a compute task on behalf of the wireless device, determining a first computing resource, of a plurality of computing resources, required to perform the compute task, associating the request with a first level of priority, of a plurality of levels of priority, for accessing the first computing resource, allocating, based at least in part on the first level of priority, a portion of the first computing resource to perform the requested compute task.

Abstract: An aspect of the disclosure includes a method, a system and a computer program product for removing dynamic images when determining the position of an object relative to a reflective surface. The method includes acquiring a first image of an object and the reflective surface with a camera, the camera positioned adjacent the reflective surface. It is determined when a dynamic element is displayed on the reflective surface. A second image is generated by subtracting the dynamic element from the first image. The object is identified in one of the first image and second image. A reflection of the object on the reflective surface is identified in the second image. Three-dimensional coordinates of at least a portion of the object are determined based at least in part on the identifying of the object and the identifying of the reflection of the object.

Abstract: The present disclosure relates to caching content at a device in a group of co-located devices whose owners share social network connections. In one embodiment, a method generally includes electing the device as a leader device based, at least in part, on shared social network connections associated with owners of each device in the group. The device receives a data request from a first user device and searches a cache for the data. If the data exists in the cache, the data is encrypted using an encryption key associated with an owner of the first user device transmitted to the first user device. Otherwise, the device requests the data from the content provider, receives an encrypted copy of the data, decrypts the data, and uses the encryption key associated with the owner of the first device to generate and transmit an encrypted message including the data to the first device.

Abstract: A computer-implemented method, according to one embodiment, includes calculating or receiving a normalized score for each segment of at least one route based on physical route conditions and environmental route conditions of the associated route, generating a pace comparison of multiple actors participating in a physical activity under different route conditions, and outputting the pace comparison. The pace comparison is based on the normalized scores of the segments of the route associated with each respective actor and a pace of each respective actor on the respective route. A computer program product for comparing paces of multiple actors, according to another embodiment, includes a computer readable storage medium having program instructions embodied therewith. The program instructions are readable and/or executable by a computer to cause the computer to perform the foregoing method.

Abstract: The present disclosure relates to processing tasks offloaded from one or more user devices to a cloud service. An example method generally includes receiving a command to be processed by the cloud service from a first device of a group of user devices. The cloud service generates a response based on the command. The cloud service determines weighting scores for each device in the group of user devices for each of audio output, video output, and haptic feedback based on context information about the group of user devices. The cloud service selects a plurality of user devices to receive at least part of the response based on the determined weighting scores, partitions the response into portions corresponding to the selected plurality of user devices, and instructs each of the selected plurality of user devices to output the one or more portions corresponding to the selected plurality of user devices.