Abstract: Methods and systems for training a machine learning model include training a machine learning model using training data. A status of the machine learning model's training is determined based on an accuracy curve of the machine learning model over the course of the training. Parameters of the training are adjusted based on the status. Training of the machine learning model is completed using the adjusted parameters.

Abstract: In one aspect, a method for operating a virtual agent will be described. An interaction context is obtained. An agent state is determined based on the obtained interaction context. The agent state indicates an activity of the virtual agent. The emotion of the virtual agent is updated based on the obtained interaction context, the determined agent state and/or a personality of the virtual agent. One or more behaviors are selected. Each behavior involves or indicates a change in an appearance of the virtual agent or generation of audio. Various embodiments relate to devices, servers, software and systems arranged to implement one or more of the above operations.

Abstract: Embodiments of the present disclosure disclose a water surface detection method, an unmanned aerial vehicle landing method and an unmanned aerial vehicle. The detection method includes: controlling the unmanned aerial vehicle to generate an airflow, where the airflow is used to generate a water surface ripple; obtaining an image of a landing area of the unmanned aerial vehicle and a flight parameter of the unmanned aerial vehicle; and processing the image, and determining, with reference to the flight parameter, whether the landing area has the water surface ripple. Damage caused by erroneous landing of the unmanned aerial vehicle on a water surface can be alleviated by determining whether the landing area has the water surface.

Abstract: An apparatus for determining a spatial disposition of a movable object includes a support base configured to be carried by the movable object, a first inertial sensor coupled to the support base via a first damping element configured to damp motion of the first inertial sensor, and a second inertial sensor coupled to the support base via a second damping element configured to damp motion of the second inertial sensor. The first inertial sensor and the second inertial sensor are of different sensor types. The first damping element and the second damping element have different damping properties.

Abstract: Various devices, arrangements and methods for managing communications using a head mounted display device are described. In one aspect, tracking data is generated at least in part by one or more sensors in a head mounted display (HMD) device. The tracking data indicates one or more facial movements of a user wearing the HMD device. A patch image is obtained based on the tracking data. The patch image is merged with a facial image. Various embodiments relate to the HMD device and other methods for generating and using the patch and facial images.

Abstract: A method of managing a multi-player poker game to be implemented using a server device communicating with a plurality of client devices each associated with a player, the method including: dealing, to each of the players, a hand of cards including a predetermined number of playing cards and a wild card that can be used to represent any one of the poker cards; providing a game rule dictating that each of the players arranges the hand of cards dealt thereto into a playing card combination; and when it is determined that each of the players has arranged the hand of cards into a playing card combination, calculating a score for each of the players according to the playing card combinations arranged respectively by the players.

Abstract: A method for dividing, by a training system, a computational training work load of one or more neural network layers; pre-training the one or more neural network layers with a first class of image data sensitive to an original known dataset; generating a first weight file from the first layer of the neural network based on the first class of image data sensitive to the original known dataset; loading the one or more pre-trained neural network layers and the generated first weight file into at least one Internet of Things (IoT) device; stacking the one or more pre-trained neural network layers with the first layer of the neural network to form a new training system for an uploaded new dataset; adjusting the generated first weight file based on an input of one or more new classes of image data comprised in the uploaded new dataset to generate a new second weight file; inferencing an object class of new image data comprised on the uploaded new dataset using the generated new second weight file; and outputting th

Abstract: The present disclosure includes methods and devices for a memory controller. In one or more embodiments, a memory controller includes a plurality of back end channels, and a command queue communicatively coupled to the plurality of back end channels. The command queue is configured to hold host commands received from a host. Circuitry is configured to generate a number of back end commands at least in response to a number of the host commands in the command queue, and distribute the number of back end commands to a number of the plurality of back end channels.

Abstract: The present invention provides a method for generating hydroxyl radicals and removing organic pollutants in water by utilizing an organic membrane. The inventive method generates hydroxyl radicals by adopting a nitrocellulose membrane as a photocatalytic material, placing the nitrocellulose membrane in water, illuminating with the sun or a sunlamp having a wavelength above 280 nm as a light source, and causing the nitrocellulose membrane to undergo a photochemical reaction at the membrane surface, and the generation rate of hydroxyl radicals can be achieved by adjusting the surface area of the membrane and a light intensity. The generated hydroxyl radicals can remove phenol, bisphenol A, thiamphenicol and other typical organic pollutants in water very well.

Abstract: The present invention provides a target detection method and an unmanned aerial vehicle. The method includes: obtaining an image of a target, and obtaining a feature model of the target according to the image of the target; determining whether the target is lost; and if it is determined that the target is lost: obtaining a currently photographed first frame of image; obtaining an Nth frame of image which is photographed after the first frame of image, N being an integer greater than 0; obtaining, according to the first frame of image and the Nth frame of image, a candidate area which possibly includes the target; and determining, according to the feature model of the target and the candidate area, whether the Nth frame of image includes the target. A detection efficiency is high by using the target detection method provided in the present invention.

Abstract: A fiber includes an elongated member and refractive index contrast domains dispersed within the elongated member. The elongated member includes at least one polymer having a transmittance of infrared radiation at a wavelength of 9.5??? of at least about 40%.

Abstract: The present invention discloses a method and an apparatus for dodging an obstacle, and a movable object. The method includes: obtaining acceleration information of a movable object; determining, according to the acceleration information of the movable object that the movable object collides; determining, according to the acceleration information, a direction for dodging an obstacle; and controlling the movable object to move along the direction for dodging the obstacle, to avoid the obstacle. In embodiments of the present invention, collision of the movable object and a dodging direction can be accurately informed according to the acceleration information of the movable object, to effectively avoid an obstacle and avoid continuously hitting the obstacle.

Abstract: Embodiments of the present disclosure disclose a water surface detection method, an unmanned aerial vehicle landing method and an unmanned aerial vehicle. The detection method includes: controlling the unmanned aerial vehicle to generate an airflow, where the airflow is used to generate a water surface ripple; obtaining an image of a landing area of the unmanned aerial vehicle and a flight parameter of the unmanned aerial vehicle; and processing the image, and determining, with reference to the flight parameter, whether the landing area has the water surface ripple. Damage caused by erroneous landing of the unmanned aerial vehicle on a water surface can be alleviated by determining whether the landing area has the water surface.

Abstract: The present invention discloses a landing protection method and apparatus and an aircraft. The method includes: obtaining an image of a landing area; determining a feature point in the image; determining, according to the feature point, whether the landing area is a dangerous landing area; and if yes, controlling the aircraft to suspend landing or controlling the aircraft to fly away from the dangerous landing area. By means of the present invention, whether a landing area is a dangerous landing area may be determined according to an obtained image, so that the landing safety of the aircraft is ensured.

Abstract: The present disclosure includes methods and devices for a memory controller. In one or more embodiments, a memory controller includes a plurality of back end channels, and a command queue communicatively coupled to the plurality of back end channels. The command queue is configured to hold host commands received from a host. Circuitry is configured to generate a number of back end commands at least in response to a number of the host commands in the command queue, and distribute the number of back end commands to a number of the plurality of back end channels.

Abstract: An apparatus for determining a spatial disposition of a movable object includes a support base configured to be carried by the movable object, a first inertial sensor coupled to the support base via a first damping element configured to damp motion of the first inertial sensor, and a second inertial sensor coupled to the support base via a second damping element configured to damp motion of the second inertial sensor. The first inertial sensor and the second inertial sensor are of different sensor types. The first damping element and the second damping element have different damping properties.

Abstract: This specification discloses an obstacle avoidance method and device, a moveable object and a computer readable storage medium. The obstacle avoidance method includes: obtaining a depth value of a spot in an area containing an obstacle, the depth value representing a distance between the spot and the moveable object; judging, according to the depth value, whether an edge of the obstacle exists in the area; and if the edge of the obstacle exists in the area, determining a movement direction of the moveable object according to the edge, so as to avoid the obstacle. According to embodiments of the present invention, an obstacle grid map does not need to be established, and a movement track does not need to be predicted, achieving advantages of a small amount of calculation and high control accuracy.

Abstract: Embodiments of the present disclosure allow user experience of photographing to be improved. In operation, it is determined whether a picture composition of a first object and a second object needs to be adjusted based on a predefined composition rule. If the picture composition needs to be adjusted, an adjusting pattern is determined based on the predefined composition rule. Then, the adjusting pattern is provided to a user to indicate the user to adjust the picture composition based on thereon.

Abstract: Improved sensing devices are provided for determining the spatial disposition of a movable object. In one aspect, an apparatus for determining a spatial disposition of a movable object comprises: a support base configured to be carried by a movable object; one or more accelerometers coupled to the support base via a first damping element configured to damp motion of the one or more accelerometers; and one or more gyroscopes coupled to the support base via a second damping element configured to damp motion of the one or more gyroscopes, wherein an amount of motion damping provided by the first damping element is greater than an amount of motion damping provided by the second damping element.

Abstract: A method of regulating a temperature of a human body includes: (1) providing an article of clothing including a textile, wherein the textile includes at least one porous layer including a polyolefin; and (2) placing the article of clothing adjacent to the human body. The porous layer has pores having an average pore size in a range of 50 nm and 1000 nm.