Abstract: A method for preparing a fully soft self-powered vibration sensor mainly uses a laser carbonization technology to prepare a two-dimensional porous carbon electrode with an origami structure, and then transfers the two-dimensional porous carbon electrode to a three-dimensional polydimethylsiloxane (PDMS) cavity through mold transfer; Finally, a laser engraving technology is used to create microstructures on surfaces of the porous carbon electrode and a PDMS film. The sensor includes the PDMS film, a liquid metal droplet oscillator, a porous out-of-plane carbon electrode, and a 3D PDMS cavity assembled tightly from top to bottom.

Abstract: An electronic device includes: a processor; a memory; an antenna, wherein a transmitting distance of the antenna at a first transmit power is a first distance; a transmitting distance of the antenna at a second transmit power lower than the first transmit power is a second distance; wherein the first distance is greater than a preset safety distance, while the second distance is less than that. The electronic device is enabled to perform sending a first message based on a first periodicity by the antenna at the first transmit power, the first message is an encrypted ciphertext; receiving a first response message of a mobile device; sending a second message based on a second periodicity by the antenna at the second transmit power, the second message is an uncrypted plaintext; receiving a second response message of the mobile device; and executing all or a part of a specific function.

Abstract: According to one aspect, cooperative multi-goal, multi-agent, multi-stage (CM3) reinforcement learning may include training a first agent using a first policy gradient and a first critic using a first loss function to learn goals in a single-agent environment using a Markov decision process, training a number of agents based on the first policy gradient and a second policy gradient and a second critic based on the first loss function and a second loss function to learn cooperation between the agents in a multi-agent environment using a Markov game to instantiate a second agent neural network, each of the agents instantiated with the first agent neural network in a pre-trained fashion, and generating a CM3 network policy based on the first agent neural network and the second agent neural network. The CM3 network policy may be implemented in a CM3 based autonomous vehicle to facilitate autonomous driving.

Abstract: A gesture recognition method and a related apparatus are provided, to obtain a first point cloud data set by filtering an original point cloud data set collected by a radar apparatus. The first point cloud data set includes a plurality of frames of first point cloud subsets, the first point cloud subset includes a first cluster center, the first cluster center is a cluster center of a plurality of pieces of point cloud data in the first point cloud subset, a maximum horizontal distance between any two first cluster centers meets a first preset condition, and duration of the first point cloud data set meets the first preset condition. Point cloud data whose motion track does not match gesture motion can be effectively filtered out. Gesture recognition is performed by using the first point cloud data set obtained by filtering.

Abstract: Disclosed are a filament device for an illumination device, the illumination device, and a dimming method for the illumination device.

Abstract: A method for customizing a key of a foldable device, a device, and a storage medium are provided. When the foldable device is in a folded state, a touchscreen of the foldable device is divided into three regions: a primary screen, a back screen, and a secondary screen. According to the method for customizing a key, the foldable device receives a first user operation used to indicate the foldable device to add a custom key on the back screen or the secondary screen, and then obtains a physical state of the foldable device. If the foldable device is in a folded state, the custom key is added on the back screen or the secondary screen based on the first user operation.

Abstract: A server receives a search request that includes a first location and an identifier of a Bluetooth device and that is sent by a first terminal; within a first time period, determines at least one second terminal by using a first frequency and a first density in a first search range, and sends a search task that carries the identifier to the at least one second terminal; if the server does not receive, within the first time period, a message that is fed back by any one of the at least one second terminal and that indicates that the Bluetooth device is found, within a second time period after the first time period, determines at least one third terminal by using a second frequency and a second density in a second search range, and sends the search task that carries the identifier to the at least one third terminal.

Abstract: Disclosed are a filament device for an illumination device, the illumination device, and a dimming method for the illumination device.

Abstract: According to one aspect, cooperative multi-goal, multi-agent, multi-stage (CM3) reinforcement learning may include training a first agent using a first policy gradient and a first critic using a first loss function to learn goals in a single-agent environment using a Markov decision process, training a number of agents based on the first policy gradient and a second policy gradient and a second critic based on the first loss function and a second loss function to learn cooperation between the agents in a multi-agent environment using a Markov game to instantiate a second agent neural network, each of the agents instantiated with the first agent neural network in a pre-trained fashion, and generating a CM3 network policy based on the first agent neural network and the second agent neural network. The CM3 network policy may be implemented in a CM3 based autonomous vehicle to facilitate autonomous driving.