A novel disparity computation technique is presented which comprises multiple orthogonal disparity maps, generated from approximately orthogonal decomposition feature spaces, collaboratively generating a composite disparity map. Using an approximately orthogonal feature set extracted from such feature spaces produces an approximately orthogonal set of disparity maps that can be composited together to produce a final disparity map. Various methods for dimensioning scenes and are presented. One approach extracts the top and bottom vertices of a cuboid, along with the set of lines, whose intersections define such points. It then defines a unique box from these two intersections as well as the associated lines. Orthographic projection is then attempted, to recenter the box perspective. This is followed by the extraction of the three-dimensional information that is associated with the box, and finally, the dimensions of the box are computed. The same concepts can apply to hallways, rooms, and any other object.

An electronic apparatus according to the present invention, includes at least one memory and at least one processor which function as: an acquisition unit configured to acquire positional information indicating a position of an object in a captured image; a display control unit configured to perform control such that an item having a length in a first direction, which corresponds to a range in a depth direction in the image, is displayed in a display, and a graphic indicating presence of the object is displayed in association with a position corresponding to the positional information in the item; a reception unit configured to be able to receive an operation of specifying a set range which is at least part of the item; and a processing unit configured to perform predetermined processing based on the set range.

An electronic apparatus according to the present invention, includes at least one memory and at least one processor which function as: an acquisition unit configured to acquire positional information indicating a position of an object in a captured image; a display control unit configured to perform control such that an item having a length in a first direction, which corresponds to a range in a depth direction in the image, is displayed in a display, and a graphic indicating presence of the object is displayed in association with a position corresponding to the positional information in the item; a reception unit configured to be able to receive an operation of specifying a set range which is at least part of the item; and a processing unit configured to perform predetermined processing based on the set range.

A method for tracking occluded objects includes encoding locations of a plurality of objects in an environment, determining a target object, receiving a first end point corresponding to a position of the target object before occlusion behind an occlusion object, distributing a hypothesis between both sides of the occlusion object during occlusion from a subsequent frame of the sequence of frames, receiving a second end point corresponding to a position of the target object after emerging from occlusion from another subsequent frame of the sequence of frames, and determining a trajectory of the target object when occluded by the occlusion object by performing inferences using a spatio-temporal probabilistic graph based on the current frame and the subsequent frames of the sequence of frames. The trajectory of the target object when occluded is used as a learning model for future target objects that are occluded by the occlusion object.

A method for tracking occluded objects includes encoding locations of a plurality of objects in an environment, determining a target object, receiving a first end point corresponding to a position of the target object before occlusion behind an occlusion object, distributing a hypothesis between both sides of the occlusion object during occlusion from a subsequent frame of the sequence of frames, receiving a second end point corresponding to a position of the target object after emerging from occlusion from another subsequent frame of the sequence of frames, and determining a trajectory of the target object when occluded by the occlusion object by performing inferences using a spatio-temporal probabilistic graph based on the current frame and the subsequent frames of the sequence of frames. The trajectory of the target object when occluded is used as a learning model for future target objects that are occluded by the occlusion object.

In various examples, a deep neural network (DNN) is trained—using image data alone—to accurately predict distances to objects, obstacles, and/or a detected free-space boundary. The DNN may be trained with ground truth data that is generated using sensor data representative of motion of an ego-vehicle and/or sensor data from any number of depth predicting sensors—such as, without limitation, RADAR sensors, LIDAR sensors, and/or SONAR sensors. The DNN may be trained using two or more loss functions each corresponding to a particular portion of the environment that depth is predicted for, such that—in deployment—more accurate depth estimates for objects, obstacles, and/or the detected free-space boundary are computed by the DNN. In some embodiments, a sampling algorithm may be used to sample depth values corresponding to an input resolution of the DNN from a predicted depth map of the DNN at an output resolution of the DNN.

In various examples, a deep neural network (DNN) is trained—using image data alone—to accurately predict distances to objects, obstacles, and/or a detected free-space boundary. The DNN may be trained with ground truth data that is generated using sensor data representative of motion of an ego-vehicle and/or sensor data from any number of depth predicting sensors—such as, without limitation, RADAR sensors, LIDAR sensors, and/or SONAR sensors. The DNN may be trained using two or more loss functions each corresponding to a particular portion of the environment that depth is predicted for, such that—in deployment—more accurate depth estimates for objects, obstacles, and/or the detected free-space boundary are computed by the DNN. In some embodiments, a sampling algorithm may be used to sample depth values corresponding to an input resolution of the DNN from a predicted depth map of the DNN at an output resolution of the DNN.

Disclosed herein is a web-based videoconference system that allows for video avatars to navigate within a virtual environment. Various methods for efficient modeling, rendering, and shading are disclosed herein.

Disclosed herein is a web-based videoconference system that allows for video avatars to navigate within a virtual environment. Various methods for efficient modeling, rendering, and shading are disclosed herein.

A device to determine a height disparity between features of an image includes a memory including instructions and processing circuitry. The processing circuitry is configured by the instructions to obtain an image including a first repetitive feature and a second repetitive feature. The processing circuitry is further configured by the instructions to determine a distribution of pixels in a first area of the image, where the first area includes an occurrence of the repetitive features, and to determine a distribution of pixels in a second area of the image, where the second area includes another occurrence of the repetitive features. The processing circuitry is further configured by the instructions to evaluate the distribution of pixels in the first area and the distribution of pixels in the second area to determine a height difference between the first repetitive feature and the second repetitive feature.