A system, method, infrastructure, and vehicle for supporting automated valet parking are provided. The automated valet parking method may include: establishing, by a vehicle, a communication with an infrastructure facility; receiving, by the vehicle, a target position and a guide route from the infrastructure facility after the communication is established; performing, by the vehicle, an autonomous driving based on the guide route; and parking, by the vehicle, at the target position.

A robotic trolley collection system and methods for automatically collecting baggage/luggage trolleys are provided. The system includes a differential-driven mobile base; a manipulator mounted on the differential-driven mobile base for forking a trolley, having a structure same as a head portion of the trolley; a sensory and measurement assembly for providing sensing and measurement dataflow; and a main processing case for processing the sensing and measurement dataflow provided by the sensory and measurement assembly and for controlling the differential-driven mobile base, the manipulator, and the sensory and measurement assembly. The method includes localizing and mapping the robotic trolley collection system; detecting an idle trolley to be collected and estimating pose of the idle trolley; visually servoing control of the robotic trolley collection system; and issuing motion control commands to the robotic trolley collection system for automatically collecting the idle trolley.

Systems and methods for generating mapping data for an autonomous vehicle (e.g., robotic devices). The methods include obtaining three-dimensional environmental data of an environment from a distance sensor. The three-dimensional environmental data includes information relating to one or more objects in the environment. The method further includes identifying at least one planar layer of two-dimensional data from the three-dimensional environmental data to be included in mapping data based on one or more characteristics of an autonomous vehicle, generating mapping data comprising the at least one planar layer of two-dimensional data from the three-dimensional environmental data, and transmitting the mapping data to the autonomous vehicle for use during operation within the environment.

Systems and methods for map construction using a video sequence captured on a camera of a vehicle in an environment, comprising: receiving a video sequence from the camera, the video sequence including a plurality of image frames capturing a scene of the environment of the vehicle; using a neural camera model to predict a depth map and a ray surface for the plurality of image frames in the received video sequence; and constructing a map of the scene of the environment based on image data captured in the plurality of frames and depth information in the predicted depth maps.

A three-layer intelligence system architecture and an exploration robot are provided. The three-layer intelligence system architecture includes: a digital twin module for creating a virtual exploration environment and a virtual robot according to explored environment data acquired in real time by the exploration robot and robot data of the exploration robot; a virtual reality module for generating a process and a result of the virtual robot executing the control commands in the virtual exploration environment according to the virtual exploration environment, the virtual robot, and control commands of a control personnel for the exploration robot; and a man-machine fusion module for transmitting the control commands and showing the control personnel the process and the result of the virtual robot executing the control commands in the virtual exploration environment, and causing the exploration robot to execute the control commands after acquiring a feedback indicating that the control personnel confirms the control commands.

The disclosure relates to method and system for localizing an autonomous ground vehicle (AGV). In an example, the method includes receiving a line drawing corresponding to a two-dimensional (2D) camera scene captured by a camera mounted on the AGV, determining a plurality of ground-touching corner edges based on a plurality of horizontal edges and a plurality of vertical edges in the line drawing, determining a plurality of three-dimensional (3D) points corresponding to a plurality of 2D points in each of the plurality of ground-touching corner edges based on a mapping relationship between an angular orientation of a ground touching edge of an object in real-world and in camera scene and a set of intrinsic parameters of the camera, generating 2D occupancy data by plotting the plurality of 3D points in a 2D plane, and determining a location of the AGV based on the 2D occupancy data and the mapping relationship.

A method for 3D object detection is described. The method includes detecting semantic keypoints from monocular images of a video stream capturing a 3D object. The method also includes inferring a 3D bounding box of the 3D object corresponding to the detected semantic vehicle keypoints. The method further includes scoring the inferred 3D bounding box of the 3D object. The method also includes detecting the 3D object according to a final 3D bounding box generated based on the scoring of the inferred 3D bounding box.

A work machine has a backup camera that captures images of an area of a worksite behind the work machine. A controller identifies pre-defined markings in the worksite and localizes the pre-defined markings to the work machine, based on the images. A control signal generator generates control signals to automatically control the work machine based upon the localized markings.

On-floor obstacle detection using an RGB-D camera is disclosed. An obstacle on a floor is detected by receiving an image including depth channel data and RGB channel data through the RGB-D camera, estimating a ground plane corresponding to the floor based on the depth channel data, obtaining a foreground of the image corresponding to the ground plane based on the depth channel data, performing a distribution modeling on the foreground of the image based on the RGB channel data to obtain a 2D location of the obstacle, and transforming the 2D location of the obstacle into a 3D location of the obstacle based on the depth channel data.

Techniques are discussed for determining a data level for portions of data for processing. In some cases, a data level can correspond to a resolution level, a compression level, a bit rate, and the like. In the context of image data, the techniques can determine a region of first image data to be processed a high resolution and a region of second image data to be processed at a low resolution. The regions can be determined by a machine learned algorithm that is trained to output identifications of such regions. Training data may be determined by identifying differences in outputs based on the first and second image data. The image data associated with the determined regions and the determined resolutions can be processed to perform object detection, classification, segmentation, bounding box generation, and the like, thereby conserving processing, bandwidth, and/or memory resources in real time systems.