Abstract: Methods and associated systems and apparatus for generating a three-dimensional (3D) flight path for a moveable platform such as an unmanned aerial vehicle (UAV) are disclosed herein. The method includes receiving a set of 3D information associated with a virtual reality environment and receiving a plurality of virtual locations in the virtual reality environment. For individual virtual locations, the system receives a corresponding action item. The system then generates a 3D path based on at least one of the set of 3D information, the plurality of virtual locations, and the plurality of action items. The system then generates a set of images associated with the 3D path and then visually presents the same to an operator via a virtual reality device. The system enables the operator to adjust the 3D path via the virtual reality device.

Abstract: Methods and systems for generating a real-time map of a survey area are provided. The methods and systems may include or be configured to carry out the steps of determining, based on a desired map resolution, a flight path over the survey area for a movable object having at least one image capture device. The methods and systems further may include obtaining images of the survey area captured by the at least one image capture device as the movable object travels along the flight path, and processing the images to generate the real-time map of the survey area with the desired map resolution.

Abstract: A method includes simulating one or more states of a movable object by implementing a movable object model. Each simulated state is associated with simulated state data of the movable object. The method further includes determining one or more sets of simulated sensor data corresponding to the one or more simulated states respectively by implementing a plurality of sensor models, determining environment data of a simulated environment surrounding the movable object by implementing an environment model, providing the one or more sets of simulated sensor data to a movable object controller configured for generating control signals to adjust states of the movable object, and providing the simulated state data and the environment data to a vision simulator configured for visualizing operations of the movable object in the one or more simulated states.

Abstract: The present invention relates to a system for efficient large-scale data distribution in a distributed and parallel processing environment. In particular, the present invention relates to global Top-k sparsification for low bandwidth networks. The present invention verifies that gTop-k S-SGD has nearly consistent convergence performance with S-SGD and evaluates the training efficiency of gTop-k on a cluster with 32 GPU machines which are inter-connected with 1 Gbps Ethernet. The experimental results show that the present invention achieves up to 2.7-12× higher scaling efficiency than S-SGD with dense gradients, and 1.1-1.7× improvement than the existing Top-k S-SGD.

Abstract: Methods and associated systems and apparatus for generating a three-dimensional (3D) flight path for a moveable platform such as an unmanned aerial vehicle (UAV) are disclosed herein. The method includes receiving a set of 3D information associated with a virtual reality environment and receiving a plurality of virtual locations in the virtual reality environment. For individual virtual locations, the system receives a corresponding action item. The system then generates a 3D path based on at least one of the set of 3D information, the plurality of virtual locations, and the plurality of action items. The system then generates a set of images associated with the 3D path and then visually presents the same to an operator via a virtual reality device. The system enables the operator to adjust the 3D path via the virtual reality device.

Abstract: Methods and associated systems and apparatus for generating a three-dimensional (3D) flight path for a moveable platform such as an unmanned aerial vehicle (UAV) are disclosed herein. The method includes receiving a set of 3D information associated with a virtual reality environment and receiving a plurality of virtual locations in the virtual reality environment. For individual virtual locations, the system receives a corresponding action item. The system then generates a 3D path based on at least one of the set of 3D information, the plurality of virtual locations, and the plurality of action items. The system then generates a set of images associated with the 3D path and then visually presents the same to an operator via a virtual reality device. The system enables the operator to adjust the 3D path via the virtual reality device.

Abstract: The present invention relates to a system for efficient large-scale data distribution in a distributed and parallel processing environment. In particular, the present invention relates to global Top-k sparsification for low bandwidth networks. The present invention verifies that gTop-k S-SGD has nearly consistent convergence performance with S-SGD and evaluates the training efficiency of gTop-k on a cluster with 32 GPU machines which are inter-connected with 1 Gbps Ethernet. The experimental results show that the present invention achieves up to 2.7-12× higher scaling efficiency than S-SGD with dense gradients, and 1.1-1.7× improvement than the existing Top-k S-SGD.

Abstract: A method of generating a three-dimensional point cloud includes obtaining a plurality of image data sources through a plurality of sensors, and performing fusion processing according to the plurality of image data sources to obtain the three-dimensional point cloud.

Abstract: A method for controlling movement of a moving object includes, based on a target object selecting operation, a center point of a target object range on a display interface and a target-object-range adjusting pressure for adjusting the target object range, determining the target object range in the display interface according to the center point of the target object range and the target-object-range adjusting pressure, and determining an image object in the target object range as a target object tracked by the moving object.

Abstract: A vision-based positioning method includes extracting first features from a first image and second features from a second image, determining initial matching pairs according to the first features and the second features, extracting satisfying matching pairs that meet a requirement from the initial matching pairs according to an affine transformation model, and determining a position-attitude change according to the satisfying matching pairs. The first image and the second image are obtained by a vision sensor carried by an aerial vehicle. The position-attitude change indicates a change from a position-attitude of the vehicle when the vision sensor captures the first image to a position-attitude of the vehicle when the vision sensor captures the second image.

Abstract: A three-dimensional (3D) reconstruction system based on aerial photography includes an unmanned aerial vehicle (UAV), a ground station, and a cloud server. The ground station is configured to determine an aerial photography parameter for indicating an aerial photography state of the UAV based on a user operation and transmit the aerial photography parameter to the UAV. The UAV is configured to receive the aerial photography parameter transmitted by the ground station; fly based on the aerial photography parameter and control an imaging device carried by the UAV to acquire aerial images during a flight; and transmit the aerial images to the cloud server. The cloud server is configured to receive the aerial images and generate a 3D model of a target area based on the aerial images.

Abstract: A method includes simulating one or more states of a movable object by implementing a movable object model. Each simulated state is associated with simulated state data of the movable object. The method further includes determining one or more sets of simulated sensor data corresponding to the one or more simulated states respectively by implementing a plurality of sensor models, determining environment data of a simulated environment surrounding the movable object by implementing an environment model, providing the one or more sets of simulated sensor data to a movable object controller configured for generating control signals to adjust states of the movable object, and providing the simulated state data and the environment data to a vision simulator configured for visualizing operations of the movable object in the one or more simulated states.

Abstract: Systems, methods, and devices are provided for controlling a movable object using multiple sensors. In one aspect, a method for calibrating one or more extrinsic parameters of a movable object having a plurality of sensors in an initial configuration is provided. The method can comprise: detecting that the initial configuration of the plurality of sensors has been modified; receiving inertial data from at least one inertial sensor during operation of the movable object; receiving image data from at least two image sensors during the operation of the movable object; and estimating the one or more extrinsic parameters based on the inertial data and the image data in response to detecting that the initial configuration has been modified, wherein the one or more extrinsic parameters comprise spatial relationships between the plurality of sensors in the modified configuration.

Abstract: Methods and systems for generating a real-time map of a survey area are provided. The methods and systems may include or be configured to carry out the steps of determining, based on a desired map resolution, a flight path over the survey area for a movable object having at least one image capture device. The methods and systems further may include obtaining images of the survey area captured by the at least one image capture device as the movable object travels along the flight path, and processing the images to generate the real-time map of the survey area with the desired map resolution.

Abstract: A method includes obtaining state data of a movable platform, scene data of an environment within which the movable platform is simulated, and parameters of one or more imaging sensors borne on the movable platform, varying one or more of the state data, the scene data, and the parameters to obtain varied data, and simulating visual data captured by the one or more imaging sensors based on the varied data for the one or more of state data, the scene data, and the parameters.

Abstract: A method includes obtaining previous state information of a movable object, receiving inertial data from at least one inertial sensor carried by the movable object, receiving image data from at least two image sensors carried by the movable object, and estimating updated state information of the movable object based on at least one of the previous state information, the inertial data, or the image data.

Abstract: A method for 3-dimensional point cloud reconstruction includes obtaining image data in a current view angle, and generating a target voxel of a target object in the current view angle according to the image data. The target voxel contains depth information. The method further includes, if a value of the depth information of the target voxel is not within a preset range, discarding the target voxel and, if the value of the depth information of the target voxel is within the preset range, fusing the target voxel with one or more stored voxels.

Abstract: Methods and associated systems and apparatus for generating a three-dimensional (3D) flight path for a moveable platform such as an unmanned aerial vehicle (UAV) are disclosed herein. The method includes receiving a set of 3D information associated with a virtual reality environment and receiving a plurality of virtual locations in the virtual reality environment. For individual virtual locations, the system receives a corresponding action item. The system then generates a 3D path based on at least one of the set of 3D information, the plurality of virtual locations, and the plurality of action items. The system then generates a set of images associated with the 3D path and then visually presents the same to an operator via a virtual reality device. The system enables the operator to adjust the 3D path via the virtual reality device.

Abstract: A visual simulation system includes a non-transitory computer-readable memory that stores computer-executable instructions and one or more processors individually or collectively configured to access the memory. The one or more processors are individually or collectively configured to execute the computer-executable instructions to generate visual data of a virtual environment obtained by a simulated visual sensor associated with a simulated movable object, obtain one or more parameters of a physical visual sensor corresponding to the simulated visual sensor, the one or more parameters comprising a noise-related parameter, and transform the generated visual data to simulate an effect of the one or more parameters.

Abstract: A method for testing an object to be tested includes obtaining a planned parameter corresponding to the object to be tested. The method also includes obtaining an actual parameter corresponding to the object to be tested through a simulation platform. The method further includes obtaining a testing result corresponding to the object to be tested based on the planned parameter and the actual parameter.