Provided is a moving robot system including a moving robot and a charging station. The charging station includes a camera formed to capture the moving robot, a communication unit configured to communicate with the moving robot, a charging contact unit configured to charge the moving robot, and a control unit configured to control the camera to receive a preview image obtained by capturing the moving robot on the basis that the moving robot having been in contact with the charging contact unit is separated from the charging contact unit. The control unit performs different types of control on the basis of whether information indicating that the moving robot is being separated from the charging contact unit is received before the moving robot is separated from the charging contact unit.

An embodiment mobility guidance system includes a sensor provided in a mobility and configured to capture a driving video or an image to transmit the driving video or the image, a memory configured to store a danger zone image, a detector configured to compare the driving video or the image captured by the sensor with the danger zone image stored in the memory to detect an existence of a danger zone on a driving path of the mobility, and a guide configured to set a guide zone in response to the detector detecting the existence of the danger zone and a change in a speed or an acceleration of the mobility, and to provide the guide zone to a second mobility.

An embodiment mobility guidance system includes a sensor provided in a mobility and configured to capture a driving video or an image to transmit the driving video or the image, a memory configured to store a danger zone image, a detector configured to compare the driving video or the image captured by the sensor with the danger zone image stored in the memory to detect an existence of a danger zone on a driving path of the mobility, and a guide configured to set a guide zone in response to the detector detecting the existence of the danger zone and a change in a speed or an acceleration of the mobility, and to provide the guide zone to a second mobility.

An autonomous off-road vehicle, upon encountering an obstruction while navigating a route, can apply a first machine-learned model to identify the obstruction. In the event that the first machine-learned model cannot identify the obstruction, the autonomous off-road vehicle can apply a second machine-learned model configured to determine whether or not the obstruction can be ignored, for instance based on dimensions of the obstruction. If the obstruction can be ignored, the autonomous off-road vehicle can continue navigating the route. If the obstruction cannot be ignored, the autonomous off-road vehicle can modify the route, can stop, can flag the obstruction to a remote human operator, can modify an interface of a human operator to display a notification or a video feed from the vehicle, and the like.

Techniques to use a trained model to determine a yaw of an object are described. For example, a system may implement various techniques to generate multiple representations for an object in an environment. Each representation vary based on the technique and data used. An estimation component may estimate a representation from the multiple representations. The model may be implemented to output a yaw for the object using the multiple representations, the estimated representation, and/or additional information. The output yaw may be used to track an object, generate a trajectory, or otherwise control a vehicle.

Techniques to use a trained model to determine a yaw of an object are described. For example, a system may implement various techniques to generate multiple representations for an object in an environment. Each representation vary based on the technique and data used. An estimation component may estimate a representation from the multiple representations. The model may be implemented to output a yaw for the object using the multiple representations, the estimated representation, and/or additional information. The output yaw may be used to track an object, generate a trajectory, or otherwise control a vehicle.

Systems and methods described herein relate to jointly training a machine-learning-based monocular optical flow, depth, and scene flow estimator. One embodiment processes a pair of temporally adjacent monocular image frames using a first neural network structure to produce a first optical flow estimate; processes the pair of temporally adjacent monocular image frames using a second neural network structure to produce an estimated depth map and an estimated scene flow; processes the estimated depth map and the estimated scene flow using the second neural network structure to produce a second optical flow estimate; and imposes a consistency loss between the first optical flow estimate and the second optical flow estimate that minimizes a difference between the first optical flow estimate and the second optical flow estimate to improve performance of the first neural network structure in estimating optical flow and the second neural network structure in estimating depth and scene flow.

Systems and methods described herein relate to jointly training a machine-learning-based monocular optical flow, depth, and scene flow estimator. One embodiment processes a pair of temporally adjacent monocular image frames using a first neural network structure to produce a first optical flow estimate; processes the pair of temporally adjacent monocular image frames using a second neural network structure to produce an estimated depth map and an estimated scene flow; processes the estimated depth map and the estimated scene flow using the second neural network structure to produce a second optical flow estimate; and imposes a consistency loss between the first optical flow estimate and the second optical flow estimate that minimizes a difference between the first optical flow estimate and the second optical flow estimate to improve performance of the first neural network structure in estimating optical flow and the second neural network structure in estimating depth and scene flow.

Systems, methods, and computer-readable media are disclosed for autonomous tennis assistant systems. Example methods include determining, by a device, an outer boundary line of a tennis court, generating a digital representation of the tennis court using the outer boundary line, where the digital representation includes at least a portion of an out-of-bounds area adjacent to the outer boundary line, determining a first location of a first tennis ball, and causing a tennis ball retrieval robot to move to the first location to retrieve the first tennis ball, where the tennis ball retrieval robot is wirelessly connected to the device.

In certain embodiments, a method includes accessing image information for a scene in a movement path of a mobile robot. The image includes image information for each of a plurality of pixels of the scene, the image information comprising respective intensity values and respective distance values. The method includes analyzing the image information to determine whether to modify the movement path of the mobile robot. The method includes initiating, in response to determining according to the image information to modify the movement path of the mobile robot, sending of a command to a drive subsystem of the mobile robot to modify the movement path of the mobile robot.