Abstract: A speech synthesis method includes: obtaining an acoustic feature sequence of a text to be processed; processing the acoustic feature sequence by using a non-autoregressive computing model in parallel to obtain first audio information of the text to be processed, wherein the first audio information comprises audio corresponding to each segment; processing the acoustic feature sequence and the first audio information by using an autoregressive computing model to obtain a residual value corresponding to each segment; and obtaining second audio information corresponding to an i-th segment based on the first audio information corresponding to the i-th segment and the residual values corresponding to a first to an (i?1)-th segment, wherein a synthesized audio of the text to be processed comprises each of the second audio information, i=1, 2 . . . n, n is a total number of the segments.

Abstract: A navigation method for a disinfection robot includes: obtaining an initial map within a field of view of the robot; generating a disinfection grid having multiple disinfection squares on the initial map, and determining multiple disinfecting consideration points in the disinfection squares; determining a disinfection order of disinfecting targets corresponding to the disinfection squares according to distances between the robot after the robot completes disinfection of one of the targets corresponding to one of the disinfection squares and to-be-disinfected ones of the disinfection squares; according to distances between the position and the disinfecting consideration points in to-be-disinfected ones of the disinfection squares, combined with priority levels of the disinfecting consideration points in to-be-disinfected ones of the disinfection squares, determining disinfecting planning points in the disinfection squares; and performing disinfection to the targets corresponding to the disinfection squares acc

Abstract: A robot state estimation method, a computer-readable storage medium, and a legged robot are provided. The method includes: obtaining force information of a left leg of a robot and a right leg of the robot; calculating a ZMP of the robot in a world coordinate system based on the force information of the left leg and the force information of the right leg; and calculating a position of a center of mass (CoM) of the robot based on a preset linear inverted pendulum model. In this manner, a brand-new linear inverted pendulum model is constructed in advance, which uses the ZMP of the robot as a supporting point of the model, thereby fully considering the influence of the change of the position of the ZMP of the robot on the position of the CoM.

Abstract: A method for navigating a robot through a limited space includes: obtaining a planned first path for the robot to reach a target position; calculating curvature and/or normal vectors of points in the first path; according to the curvature and a preset curvature threshold, and/or according to the normal vectors and a preset normal vector change threshold, obtaining a first straight path for the robot outside the limited space by curve fitting; determining an intersection of a centerline of the limited space and the first straight path, and determining a second straight path for the robot to move through the limited space according to the centerline and the intersection; and generating a second path for the robot to move through the limited space based on the first straight path and the second straight path, and navigating the robot to move through the limited space according to the second path.

Abstract: A method for generating a talking head video includes: obtaining a text and an image containing a face of a user; determining a phoneme sequence that corresponds to the text and includes one or more phonemes; determining acoustic features corresponding to the text according to the phoneme sequence, and obtaining synthesized speech corresponding to the text according to the acoustic features; determining a first mouth movement sequence corresponding to the text according to the phoneme sequence, and determining a second mouth movement sequence corresponding to the text according to the acoustic features; creating a facial action video corresponding to the user according to the first mouth movement sequence, the second mouth movement sequence and the image; and processing the synthesized speech and the facial action video synchronously to obtain a talking head video corresponding to the user.

Abstract: A walking control method for a biped robot includes: detecting whether the biped robot is in an unbalanced state; in response to detection that the biped robot is in the unbalanced state, obtaining a predicted balance step length corresponding to the biped robot in the unbalanced state; performing a smooth transition processing on the predicted balance step length according to a current movement step length of the biped robot to obtain a desired balance step length corresponding to the predicted balance step length; determining a planned leg trajectory of the biped robot according to the desired balance step length; and controlling a current swing leg of the biped robot to move according to the planned leg trajectory.

Abstract: A method includes: performing semantic segmentation on an RGBD image to obtain a semantic label of each pixel of the image; performing reconstruction of a point cloud based on the image and mapping the semantic label of each pixel of the image into the point cloud to respectively obtain a semantic point cloud of a current frame with the semantic labels and a three-dimensional scene semantic map with the semantic labels; generating two-dimensional discrete semantic feature points for each of three-dimensional semantic point clouds in the current frame and the semantic map to obtain a corresponding two-dimensional semantic feature point image, and performing a three-dimensional semantic feature description on each feature point in the two-dimensional semantic feature point image; and performing feature matching on all feature points in the current frame and all feature points in the semantic map to obtain positioning information based on the three-dimensional semantic feature description.

Abstract: A trajectory planning method, a computer-readable storage medium, and a robot are provided. The method includes: constructing a phase variable of a trajectory planning of a robot, where the phase variable is a function of two position components of a torso of the robot on a horizontal plane; and performing, using the phase variable replacing a time variable, the trajectory planning on a swinging leg of the robot in each preset coordinate axis direction. In this manner, the robot can no longer continue to follow the established trajectory after being disturbed by the environment, but make state adjustments according to the disturbance received to offset the impact of the disturbance, thereby maintaining walking stability and avoiding the problem of early or late landing of the swinging leg.

Abstract: A method for estimating a pose of a humanoid robot includes: processing obtained pose parameters of a waist of the humanoid robot and plantar motion parameters of the humanoid robot to obtain the measured pose parameters of a center point of the waist of the humanoid robot; calculating predicted pose parameters of the center point of the waist according to the obtained pose parameters of the waist; and fusing the measured pose parameters and the predicted pose parameters to obtain estimated pose parameters of the center point of the waist.

Abstract: A method of controlling a robot includes: obtaining an inertia matrix and a slack variable of the robot, and determining a momentum equation of the robot according to the inertia matrix and the slack variable; obtaining reference joint angles corresponding to a reference action of the robot; determining an optimization objective function of the momentum equation according to a first preset weight coefficient of the slack variable and a second preset weight coefficient of the reference joint angles; and determining joint angles of the robot according to the optimization objective function, and driving the robot to move according to the joint angles of the robot.

Abstract: A robot control method, a computer-readable storage medium, and a robot are provided. The method includes: obtaining first motion data, where the first motion data is human arm end motion data collected by a virtual reality device; obtaining second motion data by mapping the first motion data to a working space of an end of a robotic arm of the robot; obtaining a state of each joint of the robotic arm of the robot, and obtaining control data of the joint by performing a quadratic programming solving on the second motion data and the state of the joint; and controlling, by a motion controller of the robot, the robotic arm of the robot to move according to the obtained control data of each joint of the robot by transmitting the control data of the joint to the motion controller, so that the control method is relatively more natural, intuitive, and flexible.

Abstract: A bidirectional energy storage device for a joint includes: a sleeve comprising an open end and an opposite, closed end; an elastic member attached to the sleeve; a sliding member slidably arranged at the open end of the sleeve, wherein opposite ends of the elastic member are respectively in contact with the closed end of the sleeve and the sliding member; a first telescopic link comprising a first end pivotally connected to the sliding member and an opposite, second end; and a second telescopic link comprising a first end pivotally connected to the sliding member and a second end. The second ends of the first telescopic link and the second telescopic link are pivotally connected to a rotating member at an end of the joint, and the first telescopic link and the second telescopic link are to extend and retract to drive the sliding member to side along the sleeve.

Abstract: A method includes: acquiring an input sentence in a first language in a current round of conversation; translating the input sentence in the first language to obtain an input sentence in a second language, according to dialogue contents in the first language and dialogue contents in the second language that have a mutual translation relationship with the dialogue contents in the first language in historical rounds of conversation; invoking a multi-round conversation generation model to parse the input sentence in the second language in the current round of conversation to generate an output sentence in the second language in the current round of conversation; translating the output sentence in the second language in the current round of conversation to obtain at least one candidate result in the first language; and determining an output sentence in the first language from the at least one candidate result in the first language.

Abstract: A biped robot control methods and a biped robot using the same as well as a computer readable storage medium are provided. The method includes: obtaining an initial distance between a centroid of a double inverted pendulum model of the biped robot and a support point of the biped robot, an initial moving speed of the centroid and an initial displacement of the centroid; calculating a measured value of a stable point of the doable inverted pendulum model based on the initial distance and the initial moving speed; calculating a control output quantity based on the initial moving speed and the measured value of the stable point; calculating a desired displacement of the centroid of the double-inverted pendulum model based on the initial moving speed, the initial displacement, and the control output quantity; and controlling the biped robot to move laterally according to the desired displacement.

Abstract: A gesture recognition method includes: acquiring a target image containing a gesture to be recognized; inputting the target image to a gesture recognition model that has a first sub-model, a second sub-model, and a third sub-model, the first sub-model is to determine a gesture category and a gesture center point, the second sub-model is to determine an offset of the gesture center point, and the third sub-model is to determine a length and a width of a bounding box for the gesture to be recognized; acquiring an output result from the gesture recognition model, the output result includes the gesture category, the gesture center point, and the offset of the gesture center point, and the length and the width of the bounding box; and determining the gesture category and a position of the bounding box of the gesture to be recognized according to the output result.

Abstract: A robot stability control method includes: obtaining a desired zero moment point (ZMP) and a fed-back actual ZMP of a robot at a current moment; based on a ZMP tracking control model, the desired ZMP and the actual ZMP, calculating a desired value of a motion state of a center of mass of the robot at the current moment, wherein the desired value of the motion state of the center of mass comprises a correction amount of the position of the center of mass; based on a spring-mass-damping-acceleration model and the desired value of the motion state of the center of mass, calculating a lead control input amount for the correction amount of the position of the center of mass; and controlling motion of the robot according to the lead control input amount and a planned value of the position of the center of mass at the current moment.

Abstract: A jumping motion control method for a biped robot includes: before feet of the biped robot leaves a support surface, estimating a motion trajectory of the biped robot that leaves the support surface according to a period of time when the biped robot stays or flips in the air; calculating a first motion angle of each joint of legs of the biped robot according to the motion trajectory and inverse kinematics; determining a constraint condition according to a motion type to which an action to be performed by the biped robot corresponds; optimizing the first motion angles according to the constraint condition to obtain a second motion angle of each joint of legs of the biped robot; and controlling a jumping motion of the biped robot according to the second motion angles.