Abstract: The present disclosure relates to a method for forecasting a motion trajectory, a computer-readable storage medium, and a computer device. The method includes: obtaining an observed past trajectory of an object; obtaining a spatial pointwise feature of each trajectory point in the observed past trajectory; obtaining a temporal pointwise feature of the trajectory point according to the spatial pointwise feature of the trajectory points within a preset observation time interval; and performing motion trajectory prediction on the object according to the spatial pointwise feature and the temporal pointwise feature of the trajectory points. The present disclosure promotes the accuracy of the motion trajectory prediction.

Abstract: A method for extracting point cloud feature includes: obtaining an original point cloud, conducting hybrid scale voxelization on the original point cloud; generating point-wise hybrid scale voxel features by feature encoding the point cloud subjected to the hybrid scale voxelization; and generating pseudo image feature maps using aggregated features and projection scale information. In this way, problems that at a single voxel scale, the inference time is longer when the voxel scale is smaller, and the larger voxel fails to capture intricate features and accurate location of smaller objects can be effectively overcome.

Abstract: An object detection network includes: a hybrid voxel feature extractor configured to acquire a raw point cloud, extract a hybrid scale voxel feature from the raw point cloud, and project the hybrid scale voxel feature to generate a pseudo-image feature map; a backbone network configured to perform a hybrid voxel scale feature fusion by using the pseudo-image feature map to generate multi-class pyramid features; and a detection head configured to predict a three-dimensional object box of a corresponding class according to the multi-class pyramid features. The object detection network can effectively solve a problem that under a single voxel scale, inference time is longer if the voxel scale is smaller, and an intricate feature cannot be captured and a smaller object cannot be accurately located if the voxel scale is larger. Different classes of 3D objects can be detected quickly and accurately in a 3D scene.

Abstract: A method for encoding a point cloud feature includes: obtaining an original point cloud, generating point-wise hybrid scale voxel features by conducting hybrid scale voxelization on the original point cloud; generating voxel-wise attention features using spatial features and voxel features of points in a voxel at each voxel scale; aggregating the hybrid scale voxel features and the attention features to obtain voxel-wise projection scale information; and mapping the voxel-wise projection scale information to pseudo image features at projection scales. In this way, points can be retained during voxelization, and a feature encoding network is guided by the attention features to be more attentive to interrelationships between points in the voxels, thereby increasing accuracy of three-dimensional object detection.

Abstract: An object detection network includes: a hybrid voxel feature extractor configured to acquire a raw point cloud, extract a hybrid scale voxel feature from the raw point cloud, and project the hybrid scale voxel feature to generate a pseudo-image feature map; a backbone network configured to perform a hybrid voxel scale feature fusion by using the pseudo-image feature map to generate multi-class pyramid features; and a detection head configured to predict a three-dimensional object box of a corresponding class according to the multi-class pyramid features. The object detection network can effectively solve a problem that under a single voxel scale, inference time is longer if the voxel scale is smaller, and an intricate feature cannot be captured and a smaller object cannot be accurately located if the voxel scale is larger. Different classes of 3D objects can be detected quickly and accurately in a 3D scene.

Abstract: A method for encoding a point cloud feature includes: obtaining an original point cloud, generating point-wise hybrid scale voxel features by conducting hybrid scale voxelization on the original point cloud; generating voxel-wise attention features using spatial features and voxel features of points in a voxel at each voxel scale; aggregating the hybrid scale voxel features and the attention features to obtain voxel-wise projection scale information; and mapping the voxel-wise projection scale information to pseudo image features at projection scales. In this way, points can be retained during voxelization, and a feature encoding network is guided by the attention features to be more attentive to interrelationships between points in the voxels, thereby increasing accuracy of three-dimensional object detection.

Abstract: A method for extracting point cloud feature includes: obtaining an original point cloud, conducting hybrid scale voxelization on the original point cloud; generating point-wise hybrid scale voxel features by feature encoding the point cloud subjected to the hybrid scale voxelization; and generating pseudo image feature maps using aggregated features and projection scale information. In this way, problems that at a single voxel scale, the inference time is longer when the voxel scale is smaller, and the larger voxel fails to capture intricate features and accurate location of smaller objects can be effectively overcome.

Abstract: A method of making an integrated depth sensor window lens, such as for an augmented reality (AR) head set, the depth sensor window lens comprising a sensor lens and an illuminator lens separated by an opaque dam. The method uses a two-shot injection molding process, a first shot comprising an optically clear polymeric material to form the sensor lens and the illuminator lens and the second shot comprising an opaque polymeric material to form the separator of the two.

Abstract: A method of making an integrated depth sensor window lens, such as for an augmented reality (AR) head set, the depth sensor window lens comprising a sensor lens and an illuminator lens separated by an opaque dam. The method uses a two-shot injection molding process, a first shot comprising an optically clear polymeric material to form the sensor lens and the illuminator lens and the second shot comprising an opaque polymeric material to form the separator of the two.

Abstract: A method of making an integrated depth sensor window lens, such as for an augmented reality (AR) head set, the depth sensor window lens comprising a sensor lens and an illuminator lens separated by an opaque dam. The method uses a two-shot injection molding process, a first shot comprising an optically clear polymeric material to form the sensor lens and the illuminator lens and the second shot comprising an opaque polymeric material to form the separator of the two.

Abstract: A flexible thermal conduit runs from a first housing portion of an electronic device to a second housing portion of the electronic device, to convey heat generated by an electronic component located in the first housing portion to a heat dissipation structure located in the second housing portion, where the second housing portion is flexibly coupled to the first housing portion, for example, by a hinge or other type of joint. The flexible conduit may include a plurality of layers of thin, flat thermally conductive material, which may be arranged to flex independently of each other in the region where the first and second housing portions are coupled.

Abstract: The present invention discloses a floating platform, wherein multiple layers of compartments are configured along the height direction of the floating platform, and the center of gravity of each layer of compartments in a full-load process and a loading and unloading process is always located on a vertical line where the whole center of gravity of the floating platform is located; the multiple annular compartments are of equal-ratio subdivision in volume: the volume ratio of every two adjacent upper and lower annular compartments is inversely proportional to the density of liquid stored in the compartments; in the practical loading process, the floating platform is always kept at a constant displacement to maintain the waterplane unchanged by adjusting crude oil or seawater loaded in different layers of compartments, and thus the floating plate always has optimal hydrodynamic performance.

Abstract: A flexible thermal conduit runs from a first housing portion of an electronic device to a second housing portion of the electronic device, to convey heat generated by an electronic component located in the first housing portion to a heat dissipation structure located in the second housing portion, where the second housing portion is flexibly coupled to the first housing portion, for example, by a hinge or other type of joint. The flexible conduit may include a plurality of layers of thin, flat thermally conductive material, which may be arranged to flex independently of each other in the region where the first and second housing portions are coupled.

Abstract: A flexible thermal conduit runs from a first housing portion of an electronic device to a second housing portion of the electronic device, to convey heat generated by an electronic component located in the first housing portion to a heat dissipation structure located in the second housing portion, where the second housing portion is flexibly coupled to the first housing portion, for example, by a hinge or other type of joint. The flexible conduit may include a plurality of layers of thin, flat thermally conductive material, which may be arranged to flex independently of each other in the region where the first and second housing portions are coupled.

Abstract: A flexible thermal conduit runs from a first housing portion of an electronic device to a second housing portion of the electronic device, to convey heat generated by an electronic component located in the first housing portion to a heat dissipation structure located in the second housing portion, where the second housing portion is flexibly coupled to the first housing portion, for example, by a hinge or other type of joint. The flexible conduit may include a plurality of layers of thin, flat thermally conductive material, which may be arranged to flex independently of each other in the region where the first and second housing portions are coupled.

Abstract: A flexible thermal conduit runs from a first housing portion of an electronic device to a second housing portion of the electronic device, to convey heat generated by an electronic component located in the first housing portion to a heat dissipation structure located in the second housing portion, where the second housing portion is flexibly coupled to the first housing portion, for example, by a hinge or other type of joint. The flexible conduit may include a plurality of layers of thin, flat thermally conductive material, which may be arranged to flex independently of each other in the region where the first and second housing portions are coupled.

Abstract: A flexible thermal conduit runs from a first housing portion of an electronic device to a second housing portion of the electronic device, to convey heat generated by an electronic component located in the first housing portion to a heat dissipation structure located in the second housing portion, where the second housing portion is flexibly coupled to the first housing portion, for example, by a hinge or other type of joint. The flexible conduit may include a plurality of layers of thin, flat thermally conductive material, which may be arranged to flex independently of each other in the region where the first and second housing portions are coupled.

Abstract: The present invention discloses a floating platform, wherein multiple layers of compartments are configured along the height direction of the floating platform, and the center of gravity of each layer of compartments in a full-load process and a loading and unloading process is always located on a vertical line where the whole center of gravity of the floating platform is located; the multiple annular compartments are of equal-ratio subdivision in volume: the volume ratio of every two adjacent upper and lower annular compartments is inversely proportional to the density of liquid stored in the compartments; in the practical loading process, the floating platform is always kept at a constant displacement to maintain the waterplane unchanged by adjusting crude oil or seawater loaded in different layers of compartments, and thus the floating plate always has optimal hydrodynamic performance.

Abstract: A processing system for depositing a plurality of source materials on a substrate, includes a first thermal evaporation source that can evaporate a first source material to produce a first vapor, a second thermal evaporation source that can evaporate a second source material to produce a second vapor, a vapor mixing chamber that allows the first vapor and the second vapor to be mixed to produce a mixed vapor, and conduits that can separately transport the first vapor and the second vapor to the vapor mixing chamber. The mixed vapor can be directed toward a substrate to deposit a mixture of the first source material and the second source material on the substrate. The processing system can also include vapor filters configured to regulate flows of the first vapor and the second vapor, and a mixed vapor filter to regulate flow of the mixed vapor.