Abstract: A method comprising receiving a plurality of images of a scene captured by at least one drone; identifying features within the plurality of images; identifying similar images of the plurality of images based on the features identified within the plurality of images; comparing the similar images based on the features identified within the similar images to determine a proportion of features shared by the similar images; selecting a subset of the plurality of images that have a proportion of shared features that meets a predetermined range; generating a first 3D model of the scene from the subset of images using a first 3D model building algorithm; generating a second 3D model of the scene from the subset of images using a second 3D model building algorithm; computing errors for the first and second 3D models; and selecting as the model of the scene the first or second 3D model.

Abstract: A method comprising receiving a plurality of images of a scene captured by at least one drone; identifying features within the plurality of images; identifying similar images of the plurality of images based on the features identified within the plurality of images; comparing the similar images based on the features identified within the similar images to determine a proportion of features shared by the similar images; selecting a subset of the plurality of images that have a proportion of shared features that meets a predetermined range; generating a first 3D model of the scene from the subset of images using a first 3D model building algorithm; generating a second 3D model of the scene from the subset of images using a second 3D model building algorithm; computing errors for the first and second 3D models; and selecting as the model of the scene the first or second 3D model.

Abstract: A data processing apparatus for determining description data for describing content includes: a video captioning model to receive an input comprising at least video images associated with the content, wherein the video captioning model is trained to detect one or more predetermined motions of one or more animated objects in the video images and determine one or more captions in dependence on one or more of the predetermined motions, one or more of the captions comprising respective caption data comprising one or more words for describing one or more of the predetermined motions, the respective caption data comprising one or more of audio data, text data and image data; and output circuitry to output description data in dependence on one or more of the captions.

Abstract: Methods, systems, apparatuses, and computer program products are provided herein for determining the handedness of input provided by a user via a touch interface. For instance, for each touch-based input detected, a score indicating a probability whether the touch-based input was inputted by a particular hand of the user is generated. A classification for the touch-based input is then generated based on a drift diffusion model-based technique in which inter-dependencies between a series of touch-based input are approximated. The determined classifications are used to determine the handedness of the user.

Abstract: Methods, systems, apparatuses, and computer program products are provided herein for determining the handedness of input provided by a user via a touch interface. For instance, for each touch-based input detected, a score indicating a probability whether the touch-based input was inputted by a particular hand of the user is generated. A classification for the touch-based input is then generated based on a drift diffusion model-based technique in which inter-dependencies between a series of touch-based input are approximated. The determined classifications are used to determine the handedness of the user.

Abstract: Embodiments described herein are directed to training techniques to reduce the power consumption and decrease the inference time of an NN. For example, during training, an estimate of power consumed by AMACs of a hardware accelerator on which the NN executes during inferencing is determined. The estimate is based at least on the non-zero midterms generated by the AMACs and the precision thereof. A loss function of the NN is modified such that it formulates the non-zero midterms and the precision thereof. The training forces the modified loss function to generate a sparse bit representation of the weights of the NN and to reduce the precision of the AMACs. Noise may also be injected at the output of nodes of the NN that emulates noise generated at an output of the AMACs. This enables the weights to account for the intrinsic noise that is experienced by the AMACs during inference.

Abstract: Methods, systems, apparatuses, and computer program products are provided herein for determining the handedness of input provided by a user via a touch interface. For instance, for each touch-based input detected, a score indicating a probability whether the touch-based input was inputted by a particular hand of the user is generated. A classification for the touch-based input is then generated based on a drift diffusion model-based technique in which inter-dependencies between a series of touch-based input are approximated. The determined classifications are used to determine the handedness of the user.

Abstract: A central command system may determine a mission plan for resilient execution by a swarm of drones comprising one or more sensors to capture data in accordance with the mission plan. The mission plan may specify requirements for fault tolerance or parallelism and a redundancy structure for the swarm. The mission plan may be transmitted to a remote drone swarm controller device that determines a swarm configuration based on the mission plan and available drones. The controller may transmit instructions regarding the swarm configuration to dispatch a resilient swarm of drones. During execution of the mission plan, drones in the resilient swarm may be monitored by other drones in the swarm, by the remote drone swarm controller, and/or by the central command system. The redundancy structure provides for failover options for one or more drones in the resilient swarm.

Abstract: Systems and methods for optimal selection of beats for a 3D mapping are disclosed. A method in accordance with the present disclosure may be performed on a processor and may comprise receiving a plurality of beats from a catheter. The catheter may be located at a target mapping site, such as a chamber of the heart. A plurality of dynamic filters may be applied to the plurality of collected beats. The optimal beats may be determined and integrated as a beat in the 3D mapping system. This method enables the selection of more beats for a more comprehensive 3D mapping, as well as the selection of more quality beats that are representative of the target anatomy.

Abstract: Methods, systems, apparatuses, and computer program products are provided herein for determining the handedness of input provided by a user via a touch interface. For instance, for each touch-based input detected, a score indicating a probability whether the touch-based input was inputted by a particular hand of the user is generated. A classification for the touch-based input is then generated based on a drift diffusion model-based technique in which inter-dependencies between a series of touch-based input are approximated. The determined classifications are aggregated and compared to threshold(s) to determine the handedness of the user. For example, if the aggregated classifications meet a first threshold, then a determination is made that touch-based input provided by a user was inputted by the user's left hand. If the aggregated classifications meet a second threshold, then a determination is made that touch-based input provided by the user was inputted by the user's right hand.

Abstract: A method comprising receiving a plurality of images of a scene captured by at least one drone; identifying features within the plurality of images; identifying similar images of the plurality of images based on the features identified within the plurality of images; comparing the similar images based on the features identified within the similar images to determine a proportion of features shared by the similar images; selecting a subset of the plurality of images that have a proportion of shared features that meets a predetermined range; generating a first 3D model of the scene from the subset of images using a first 3D model building algorithm; generating a second 3D model of the scene from the subset of images using a second 3D model building algorithm; computing errors for the first and second 3D models; and selecting as the model of the scene the first or second 3D model.

Abstract: A method for improving wireless communication for a drone swarm, the method comprising, at a computing system, receiving, from a plurality of drones of a drone swarm, data comprising radio frequency signal characteristics detected by the plurality of drones; generating a model of a radio frequency environment for the drone swarm based on the data received from the plurality of drones; and controlling at least one wireless communication system to improve wireless communication for the drone swarm based on the model of the radio frequency environment.

Abstract: A method for controlling a plurality of drones to survey a location, the method comprising, at a computing system: automatically generating preliminary flight plans for a plurality of drones to survey the location based on a 3D model; receiving survey data from the plurality of drones as the plurality of drones are surveying the location based on the preliminary flight plans; updating the 3D model based on the survey data received from the plurality of drones; and automatically updating at least a portion of the flight plans based on the updated 3D model

Abstract: The present invention relates to a face mask comprising (a) a mask body portion configured to surround the mouth and at least part of the nose of a wearer; (b) a sealed compressible cushion formed on the inner face of said mask body portion wherein the middle of said cushion is placed in close proximity to the middle of the superior perimeter of said mask body portion, and wherein said cushion is at least partially filled with a fluid.

Abstract: A method is provided. The method includes receiving input intracardiac signals from a monitoring and processing apparatus. Each of the input intracardiac signals includes artifacts. The method includes encoding, by an autoencoder, the input intracardiac signals utilizing an intracardiac dataset to produce a latent representation. The method also includes decoding, by an autoencoder, the latent representation to produce output intracardiac signals. The output intracardiac signals include the input intracardiac signals reconstructed without the signal artifacts.

Abstract: A mass data storage system includes a number of communicatively coupled storage drives powered by one or more power supplies. Shared control electronics selectively connect power and a data signal to a select storage drive via instructions within a control signal received by the common controller. Instructions for selectively powering and connecting the data signal are transmitted over a first signal path to a first controller of the shared electronics. Responsive to successful execution of the instructions, a drive access command is sent over a second different signal path to a second controller of the shared electronics.

Abstract: A mass data storage system includes a plurality of communicatively coupled storage drives powered by one or more power supplies. A power map defines the relationships between the storage drives and the power supplies and power rules/policies define the maximum permissible power load on each power supply at any point in time.

Abstract: A mass data storage system includes a data manager that selects a subset of storage resources for storage of a data file and generates location metadata for the data file defining positions of each storage resource in the selected subset of storage resources. According to one implementation, the data manager further defines an extended file attribute associated with the location metadata.