Techniques related to implementing fully convolutional networks for semantic image segmentation are discussed. Such techniques may include combining feature maps from multiple stages of a multi-stage fully convolutional network to generate a hyper-feature corresponding to an input image, up-sampling the hyper-feature and summing it with a feature map of a previous stage to provide a final set of features, and classifying the final set of features to provide semantic image segmentation of the input image.

Techniques are disclosed for measuring the corpus callosum volume of a fetus using magnetic resonance imaging. A scanogram of a fetus is acquired, and a detection area is determined using the corpus callosum position of the fetus in the scanogram. Magnetic resonance scanning is performed on the detection area to obtain a diffusion weighted image, with a gradient direction that is orthogonal or normal to an extending direction of fiber bundles of the corpus callosum. A fetal head image is cropped in the diffusion weighted image, and a predetermined threshold is applied to obtain an image including pixels having a brightness value that is greater than the threshold. Image processing is performed on the binarized image, with the largest region therein being identified as the corpus callosum, and the sum of voxel dimensions associated with the signal of the largest region being calculated as the corpus callosum volume.

An orientation tag provides orientation information and, by affixing the orientation tag to an object, orientation information of the object. The orientation tag displays different orientation information based on viewing the orientation tag at different angles. In particular, the orientation tag displays an angle and direction that the orientation tag is rotated about a horizontal axis relative to a viewer, and an angle and direction that the orientation tag is rotated about a vertical axis relative to the viewer. Viewing the orientation tag enables determining an angle and direction (e.g., depth information) that the orientation tag is rotated about a depth axis relative to the viewer. The orientation information and the depth information facilitate determining the orientation of the orientation tag in three dimensions. An output device outputs a user interactive experience based on the orientation information and the depth information provided by the orientation tag.

An orientation tag provides orientation information and, by affixing the orientation tag to an object, orientation information of the object. The orientation tag displays different orientation information based on viewing the orientation tag at different angles. In particular, the orientation tag displays an angle and direction that the orientation tag is rotated about a horizontal axis relative to a viewer, and an angle and direction that the orientation tag is rotated about a vertical axis relative to the viewer. Viewing the orientation tag enables determining an angle and direction (e.g., depth information) that the orientation tag is rotated about a depth axis relative to the viewer. The orientation information and the depth information facilitate determining the orientation of the orientation tag in three dimensions. An output device outputs a user interactive experience based on the orientation information and the depth information provided by the orientation tag.

In various examples, sensor data may be adjusted to represent a virtual field of view different from an actual field of view of the sensor, and the sensor data—with or without virtual adjustment—may be applied to a stereographic projection algorithm to generate a projected image. The projected image may then be applied to a machine learning model—such as a deep neural network (DNN)—to detect and/or classify features or objects represented therein.

A data labeling method, apparatus and system are provided. The method includes: sampling a data source according to an evaluation task for the data source to obtain sampled data; generating a labeling task from the sampled data; sending the labeling task to a labeling device; and receiving a labeled result of the labeling task from the labeling device. As such, an automatic evaluation of data can be implemented by using the evaluation task, and evaluation efficiency is improved.

A landslide recognition method based on Laplacian pyramid remote sensing image fusion includes: performing original remote sensing image reconstruction based on extracted local features and global features of remote sensing images through a Laplacian pyramid fusion module to generate a fused image, constructing a deep learning semantic segmentation model through a semantic segmentation network, labeling the fused image to obtain a dataset of landslide disaster label map, and training the deep learning semantic segmentation model by the dataset, and then storing when a loss curve is fitted and a landslide recognition accuracy of remote sensing image of the deep learning semantics segmentation model meets a requirement by modifying a structure of the semantic segmentation network and adjusting parameters of the deep learning semantics segmentation model. Combined with the image fusion model based on Laplacian pyramid, the method can provide effective decision-making basis for prevention and mitigation of landslide disasters.

A safety-based prediction apparatus, system and method are provided. A machine learning hardware accelerator (MLHA) includes a main classifier (MC) module, at least one guardian classifier (GC) module, and a final predicted class decision module. The MC module predicts an MC predicted class based on input data, and includes a pre-trained, machine learning main classifier (MLMC) that has at least one safety critical (SC) class and a plurality of non-SC classes. Each guardian classifier (GC) module is associated with an SC class, and predicts a GC predicted class based on the input data. Each GC module includes a pre-trained, machine learning guardian classifier (MLGC) having two classes including an associated SC class and a residual class that includes any non-associated SC classes and the plurality of non-SC classes. A decision module determines and outputs a final predicted class based on the MC predicted class and each GC predicted class.

Elements to be provided for display with an instance of content can be determined dynamically based upon features of those elements. The actions or behavior of a user can be stored over time, and used to determine element features that are of interest to the user for various categories of content. In order to obtain enough data for multiple categories, clustering of users can be performed where users with similar behaviors are grouped together, and their preferences for features determined for various categories. When a request for content is received, a category and user identity can be determined for the request, which is used to determine the appropriate cluster. The feature preferences for that cluster can then be used to select elements (e.g., images, audio, or video) to present with the content based at least in part upon the relevant features of those elements.

A method of an escape reorder mode for neural network model compression, is performed by at least one processor, and includes determining whether a frequency count of a codebook index included in a predicted codebook is less than a predetermined value, the codebook index corresponding to a neural network. The method further includes, based on the frequency count of the codebook index being determined to be greater than the predetermined value, maintaining the codebook index, and based on the frequency count of the codebook index being determined to be less than the predetermined value, assigning the codebook index to be an escape index of 0 or a predetermined number. The method further includes encoding the codebook index, and transmitting the encoded codebook index.