Abstract: A pet litter device includes: a base configured to be placed on a support surface, the base defining a recessed portion that comprises a recess surface defining an opening, and a drum rotatably connected to the base and partly received in the recessed portion. The drum has an outer lateral surface. The recess surface and a portion of the outer lateral surface that faces the recess surface define a gap and form a passage in communication with the opening. The recess surface is to guide litter particles entering the passage to pass through the opening to fall on the support surface.

Abstract: A collision detection method, a storage medium, and a robot are provided. The method includes: calculating an external torque of a first joint of the robot based on a preset generalized momentum-based disturbance observer; calculating an external torque of a second joint of the robot based on a preset long short-term memory network; calculating an external torque of a third joint of the robot based on the external torque of the first joint and the external torque of the second joint; and determining whether the robot has collided with an external environment or not based on the external torque of the third joint and a preset collision threshold. In the present disclosure, the component of the model error in the joint external torque calculated by the disturbance observer is eliminated to obtain the accurate contact torque, thereby improving the accuracy of the collision detection.

Abstract: A method for controlling a legged robot includes: in response to detection of a collision event associated with a foot of a swing leg of the biped robot, terminating a trajectory component planning of the swing leg in a collision direction; calculating a position offset in the collision direction according to an external force that is received by the foot of the swing leg in the collision direction and obtained in real time, based on a foot dragging control mode, and determining a replanned trajectory component in the collision direction based on the position offset; and controlling the swing leg to move based on the replanned trajectory component in the collision direction and a desired trajectory component of the swing leg in a non-collision direction.

Abstract: A robot balance control method, a controller, and a computer readable storage medium are provided. The method includes: obtaining a desired motion trajectory matching a current motion status by performing a parameter adaptation adjustment on a current planned motion trajectory; determining, according to the motion status, a desired state parameter of each of soles, a centroid, and a waist of a humanoid robot for conforming to the desired motion trajectory; calculating, based on the motion status and the desired state parameter of each of the soles, the centroid, and the waist of the humanoid robot, a desired driving parameter of the humanoid robot for simultaneously meeting a robot dynamics requirement, a sole control requirement, a centroid control requirement, a waist control requirement, and force control parameter distribution constraint(s) at the current moment; and controlling, based on the desired driving parameter, a movement of the humanoid robot.

Abstract: A stepping down trajectory planning method as well as a robot using the same and a computer readable storage medium are provided. The method includes: dividing a stepping down process of the robot into a plurality of planned stages; adjusting a start position of a swing leg of the robot according to an ankle-to-heel distance, where the ankle-to-heel distance is a horizontal distance between an ankle joint of the swing leg of the robot and a heel of the swing leg of the robot; determining an initial state and an end state of the swing leg in each of the planned stages according to the start position; and obtaining a planned trajectory of the swing leg by performing a curve fitting on the swing leg in each of the planned stages the initial state and the end state.

Abstract: A robot obstacle avoidance method, a robot controller using the same, and a storage medium are provided. The method includes: determining an influence value of an obstacle on a motion range of a joint of the robot according to a position of the obstacle in a workspace of the robot; establishing a state transition relationship of the robot by taking a joint velocity of the robot as a control target and a joint angular velocity of the robot as a control input quantity; and avoiding the robot from colliding with the obstacle during a movement process of the robot by performing a model predictive control on the robot according to the state transition relationship and the influence value. In the present disclosure, the influence of the obstacle on the motion range of the joint of the robot is fully considered.

Abstract: A path planning method and a biped robot using the same are provided. The method includes: generating a candidate node set for a next foot placement based on a biped robot's own parameters and joint information of a current node, adding valid candidate nodes in the candidate node set to a priority queue so as to select optimal nodes for realizing next node expansion. These optimal nodes are output to generate a foot placement sequence from an initial node to a target node, which can greatly reduce the search amount for path nodes when the robot's legs intersect and touch the ground, thereby improving the efficiency of path planning.

Abstract: A decoupling control method for a humanoid robot includes: decomposing tasks of the humanoid robot to obtain kinematic tasks and dynamic tasks, and classifying corresponding joints of the humanoid robot into kinematic task joints or dynamic task joints; solving desired positions and desired speeds of the kinematic task joints for performing the kinematic tasks according to desired positions and desired speeds of ends in the kinematic tasks using inverse kinematics; calculating torques of the kinematic task joints based on the desired positions and desired speeds of the kinematic task joints; and solving a pre-built optimization model of torques required for the dynamic task joints based on the calculated torques of the kinematic task joints, to obtain torques required by the dynamic task joints for performing the dynamic tasks.

Abstract: A method for generating a center of mass (CoM) trajectory includes determining an actual pose of a center of mass (CoM), a pose of a left foot, and a pose of a right pose of a robot; determining a first pose tracking vector of the robot according to the actual pose of the CoM and the pose of the left foot, and determining a second pose tracking vector of the robot according to the actual pose of the CoM and the pose of the right foot; and controlling a desired pose of the CoM of the robot to alternately track the pose of the left foot and the pose of the right foot, according to the first pose tracking vector and the second pose tracking vector, so as to generate a desired CoM trajectory of the robot.

Abstract: Image processing methods are provided. One of the method includes: obtaining a to-be-processed multi-channel feature maps; obtaining multi-channel first output feature maps and multi-channel second output feature maps by processing the multi-channel feature maps through a parallel pointwise convolution and non-pointwise operation, where the non-pointwise convolution is for descripting a spatial feature of each channel and an information exchange between the feature maps; and fusing the multi-channel first output feature maps and the multi-channel second output feature maps to obtain a multi-channel third output feature map.

Abstract: A control method for a robot includes: determining a desired zero moment point (ZMP) of the robot; obtaining a position of a left foot and a position of a right foot of the robot, and calculating desired support forces of the left foot and the right foot according to the desired ZMP, the positions of the left foot and the right foot; obtaining measured support forces of the left foot and the right foot, and calculating an amount of change in length of the left leg and an amount of change in length of the right leg according to the desired support forces of the left foot and the right foot, the measured support forces of the left foot and the right foot; and controlling the robot to walk according to the amount of change in length of the left leg and the right leg.

Abstract: A method for training a multi-task recognition model includes: obtaining a number of sample images, wherein some of the sample images are to provide feature-independent facial attributes, some of the sample images are to provide feature-coupled facial attributes, and some of the sample images are to provide facial attributes of face poses; training an initial feature-sharing model based on a first set of sample images to obtain a first feature-sharing model; training the first feature-sharing model based on the first set of sample images and a second set of sample images to obtain a second feature-sharing model with a loss value less than a preset second threshold; obtaining an initial multi-task recognition model by adding a feature decoupling model to the second feature-sharing model; and training the initial multi-task recognition model based on the sample images to obtain a trained multi-task recognition model.

Abstract: A robot balance control method as well as a robot using the same and a computer readable storage medium are provided. In the method, a brand new flywheel model different from the existing flywheel model is created. In this flywheel model, the foot of the support leg of the robot is equivalent to the massless link of the flywheel model, while rest parts of the robot are equivalent to the flywheel of the flywheel model. Compared with the various models in the prior art, this flywheel model is more in line with the actual situation of the robot during the monoped supporting period. By controlling the posture of the foot of the support leg based on this flywheel model, a better balance effect can be achieved, which avoids the overturning of the robot.

Abstract: A robot control method includes: determining a planned capture point and a measured capture point of the robot so as to calculate a capture point error of the robot; obtaining positions of a left foot and a right foot of the robot, and a planned zero moment point (ZMP) of the robot so as to calculate desired support forces of the left foot and the right foot; calculating desired torques of the left foot and the right foot according to the capture point error, the desired support forces of the left foot and the right foot; obtaining measured torques of the left foot and the right foot so as to calculate desired poses of the left foot and the right foot; and controlling the robot to walk according to the desired poses of the left foot and the desired pose of the right foot.

Abstract: Human lying posture detections are disclosed. A human lying on a bed is detected by obtaining an image through a depth camera, detecting objects in the image and marking the objects in the image using 2D bounding boxes by deep learning, determining the human being in a lying posture in response to a width and a height of the 2D bounding box of the human meeting a predetermined condition, detecting one or more skin areas in the image and generating skin area 2D bounding boxes to mark each of the one or more skin areas using a skin detection algorithm, and determining the human being in the lying posture in response to the skin area 2D bounding boxes and the 2D bounding box of the bed meeting a predetermined positional relationship.

Abstract: A pose calibration method, a robot, and a computer readable storage medium are provided. The method includes: obtaining, through a depth camera on a robot, a depth image including a target plane (i.e., a plane where the robot is located); determining point cloud data corresponding to the depth image; and calibrating a target pose of the depth camera based on the point cloud data and a preset optimization method, that is, calibrating a pitch angle and a roll angle of the depth camera and a height of the depth camera in a coordinate system of the robot. In this manner, the accuracy of the calibration of the target pose can be effectively improved while simple in implementation and small in calculation amount, and the efficiency of the calibration of the target pose can be improved so as to improve the user experience.

Abstract: A pose control method for a robot includes: estimating a first set of joint angular velocities of all joints of the robot according to a balance control algorithm; estimating a second set of joint angular velocities of all joints of the robot according to a momentum planning algorithm; estimating a third set of joint angular velocities of all joints of the robot according to a pose return-to-zero algorithm; and performing pose control on the robot according to the first set of joint angular velocities, the second set of joint angular velocities, and the third set of joint angular velocities.