Abstract: Calibration of eye tracking is improved by collecting additional calibration pairs when user is using apps with eye tracking. A user input component is presented on a display of an electronic device, detecting a dwelling action for user input component, and in response to detecting the dwelling action, obtaining a calibration pair comprising an uncalibrated gaze point and a screen location of the user input component, wherein the uncalibrated gaze point is determined based on an eye pose during the dwelling action. A screen gaze estimation is determine based on the uncalibrated gaze point, and in response to determining that the calibration pair is a valid calibration pair, training a calibration model using the calibration pair.

Abstract: Methods, systems, and apparatuses are described causing light to be emitted, causing a frequency at which the light is emitted to vary, receiving, based on the frequency variation, a user input, determining a critical flicker frequency (CFF) corresponding to the user input, and determining, based on the CFF, a disease state.

Abstract: A honeycomb system includes at least one modular storage unit. The modular storage unit comprises a set of dual-goods-location shelves at least one lifting machine, at least one shuttle vehicle and at least one power station. The set of dual-goods-location shelves are parallel to each other to form a roadway where the shuttle vehicle operates. The power station is arranged at two ends or in a middle portion of the dual-goods-location shelves and is integrated with a transmission mechanism and a support frame. The transmission mechanism is connected to the shuttle vehicle and the lifting machine to perform goods transfer therebetween.

Abstract: In an example embodiment, position bias and other types of bias may be compensated for by using two-phase training of a machine-learned model. In a first phase, the machine-learned model is trained using non-randomized training data. Since certain types of machine-learned models, such as those involving deep learning (e.g., neural networks) require a lot of training data, this allows the bulk of the training to be devoted to training using non-randomized training data. However, since this non-randomized training data may be biased, a second training phase is then used to revise the machine-learned model based on randomized training data to remove the bias from the machine-learned model. Since this randomized training data may be less plentiful, this allows the deep learning machine-learned model to be trained to operate in an unbiased manner without the need to generate additional randomized training data.

Abstract: This application discloses a method for implementing a software-defined network (SDN), including: receiving, by an SDN controller, virtual network information sent by a cloud computing platform, where the virtual network information includes a correspondence between information about a virtual machine (VM) and information about a server running the VM. The method also includes acquiring, by the SDN controller, physical network information, where the physical network information includes a correspondence between information about a top-of-rack (TOR) switch and the information about the server. Additionally, the method includes acquiring, by the SDN controller, a correspondence between the information about the VM and the information about the TOR switch according to the virtual network information and the physical network information.

Abstract: A method for implementing a software-defined network (SDN) includes receiving, by an SDN controller, virtual network information sent by a cloud computing platform, where the virtual network information includes a correspondence between information about a virtual machine (VM) and information about a server running the VM. The method also includes acquiring, by the SDN controller, physical network information, where the physical network information includes a correspondence between information about a top-of-rack (TOR) switch and the information about the server. Additionally, the method includes acquiring, by the SDN controller, a correspondence between the information about the VM and the information about the TOR switch according to the virtual network information and the physical network information.

Abstract: The invention provides a method of producing graphene. The method comprising: A) mixing graphite powders with a silk fibroin nanofiber solution, performing mechanical stirring to exfoliate graphite to form graphene flakes; wherein the silk fibroin nanofibers in the silk fibroin nanofiber solution have a crystallinity of 40% or above; the silk fibroin nanofibers have a diameter of 10 to 30 nm; the silk fibroin nanofibers have a length of 100 nm to 3 ?m; the mechanical stirring has a shearing speed of 1,000 to 50,000 rpm; and the duration of the mechanical stirring is 10 min to 6 h; B) centrifuging the solution obtained in step A) after exfoliation to remove unexfoliated graphite; and C) centrifuging the centrifuged solution obtained in step B), and separating graphene from the silk fibroin nanofibers to obtain the graphene.

Abstract: In one embodiment, a method includes accessing, from a data store of the communication network, user information associated with a first user of the communication network, identifying one or more entities associated with the communication network that are relevant to the first user based on the user information, and retrieving, for each identified entity, one or more content frames associated with the entity. The method includes ranking the one or more content frames based on the user information. The method also includes sending, to a client device of the first user, one or more of the content frames for display to the first user in ranked order, wherein each content frame is selectable by the first user to display the selected content frame in association with a particular content item for a specified period of time.

Abstract: Techniques for text entry using gestures are disclosed. As disclosed, a camera may capture a frame and the face of the user can be detected therein. Landmarks can be aligned to the face in the captured frame. A left eye image and a right eye image may be extracted from the captured frame. The left eye image and the right image each may be resized and compared to a calibration template. A direction of eye gaze may be determined based upon the comparison. A character or word may be predicted based upon the determination of the direction of eye gaze and a known configuration of an eye gaze board (e.g., an E-tran board). The predicted character or word can be included as a part of a text-based message.

Abstract: Techniques for text entry using gestures are disclosed. As disclosed, a camera may capture a frame and the face of the user can be detected therein. Landmarks can be aligned to the face in the captured frame. A left eye image and a right eye image may be extracted from the captured frame. The left eye image and the right image each may be resized and compared to a calibration template. A direction of eye gaze may be determined based upon the comparison. A character or word may be predicted based upon the determination of the direction of eye gaze and a known configuration of an eye gaze board (e.g., an E-tran board). The predicted character or word can be included as a part of a text-based message.

Abstract: In one embodiment, a method includes accessing, from a data store of the communication network, user information associated with a first user of the communication network, identifying one or more entities associated with the communication network that are relevant to the first user based on the user information, and retrieving, for each identified entity, one or more content frames associated with the entity. The method includes ranking the one or more content frames based on the user information. The method also includes sending, to a client device of the first user, one or more of the content frames for display to the first user in ranked order, wherein each content frame is selectable by the first user to display the selected content frame in association with a particular content item for a specified period of time.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for identifying a direction in which a user is looking. In one aspect, a method includes receiving an image of a sequence of images. The image can depict a face of a user. A template image for each particular facial feature point can be compared to one or more image portions of the image. The template image for the particular facial feature point can include a portion of a previous image of the sequence of images that depicted the facial feature point. Based on the comparison, a matching image portion of the image that matches the template image for the particular facial feature point is identified. A location of the matching image portion is identified in the image. A direction in which the user is looking is determined based on the identified location for each template image.

Abstract: This application discloses a method for implementing a software-defined network (SDN), including: receiving, by an SDN controller, virtual network information sent by a cloud computing platform, where the virtual network information includes a correspondence between information about a virtual machine (VM) and information about a server running the VM. The method also includes acquiring, by the SDN controller, physical network information, where the physical network information includes a correspondence between information about a top-of-rack (TOR) switch and the information about the server. Additionally, the method includes acquiring, by the SDN controller, a correspondence between the information about the VM and the information about the TOR switch according to the virtual network information and the physical network information.

Abstract: Embodiments of the present invention provide a method and a device for link aggregation control protocol (LACP) link switch. The method includes: monitoring state changes of interfaces in an LACP link aggregation group; when monitoring a state change occurs on an active interface in the aggregation group, determining whether the number of active interfaces in the aggregation group is smaller than a preset minimum active link number, if yes, keeping states of N interfaces among the interfaces on which a state change occurs being an active state, so that the number of active interfaces in the aggregation group is greater than or equal to the preset minimum active link number; and when monitoring a state change occurs on an inactive interface in the aggregation group, adjusting states of M interfaces among the N interfaces kept in the active state into an inactive state M being smaller than or equal to N.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for identifying a direction in which a user is looking. In one aspect, a method includes receiving an image of a sequence of images. The image can depict a face of a user. A template image for each particular facial feature point can be compared to one or more image portions of the image. The template image for the particular facial feature point can include a portion of a previous image of the sequence of images that depicted the facial feature point. Based on the comparison, a matching image portion of the image that matches the template image for the particular facial feature point is identified. A location of the matching image portion is identified in the image. A direction in which the user is looking is determined based on the identified location for each template image.

Abstract: Embodiments of the present invention provide a method and a device for link aggregation control protocol (LACP) link switch. The method includes: monitoring state changes of interfaces in an LACP link aggregation group; when monitoring a state change occurs on an active interface in the aggregation group, determining whether the number of active interfaces in the aggregation group is smaller than a preset minimum active link number, if yes, keeping states of N interfaces among the interfaces on which a state change occurs being an active state, so that the number of active interfaces in the aggregation group is greater than or equal to the preset minimum active link number; and when monitoring a state change occurs on an inactive interface in the aggregation group, adjusting states of M interfaces among the N interfaces kept in the active state into an inactive state M being smaller than or equal to N.

Abstract: Embodiments of the present invention provide a method for reducing the capacity of a Virtual Concatenation Group (VCG) and a VCG communication network. When a network management device sends a downlink-member removing command to remove a downlink member to a Sink in the VCG, the Sink receiving the downlink-member removing command in the VCG sends, to a Source, failure state information indicating that the downlink member will be removed; the Source receiving the failure state information in the VCG sends do-not-use information to the Sink indicating that the downlink member will not be used any longer, and the Sink removes the downlink member; the network management device sends an uplink-member removing command to remove an uplink member to the Source, the Source removes the uplink member which will not be used any longer. Thus, possibility of mis-operation can be avoided fundamentally and operational freedom of an operator is enhanced.

Abstract: Embodiments of the present invention provide a method for reducing the capacity of a Virtual Concatenation Group (VCG) and a VCG communication network. When a network management device sends a downlink-member removing command to remove a downlink member to a Sink in the VCG, the Sink receiving the downlink-member removing command in the VCG sends, to a Source, failure state information indicating that the downlink member will be removed; the Source receiving the failure state information in the VCG sends do-not-use information to the Sink indicating that the downlink member will not be used any longer, and the Sink removes the downlink member; the network management device sends an uplink-member removing command to remove an uplink member to the Source, the Source removes the uplink member which will not be used any longer.