A method of determining a state of a target object, an electronic device, and a storage medium, relate to fields of a computer technology, cloud computing and Internet of things, and apply to smart cities. The method includes: receiving a transmitted first moving point sequence for the target object, the first moving point sequence including a plurality of target moving point elements, and each target moving point element containing a timestamp information and a displacement information that indicate a stay state of the target object; determining, from the first moving point sequence, a target stay point of the target object, according to the timestamp information and the displacement information; and determining that the state of the target object at the target stay point is an abnormal stay state, in response to a distance between the target stay point and a first preset position being less than a first preset threshold.

Provided are a training and detection method and apparatus of an object detection network and a device and a storage medium. The method of training an object detection network includes: obtaining, by performing object detection for images in an image data set input into the object detection network and for each of one or more objects involved in each of the images, a confidence levels that the object is predicted as each of a plurality of preset categories; for each of the objects, determining reference labeling information of the object with respect to each of the non-labeled categories; for each of the objects, determining loss information that the object is predicted as each of the preset categories; and adjusting a network parameter of the object detection network based on the loss information that each of the objects is predicted as each of the preset categories.

The present invention discloses an adaptive-learning intelligent scheduling unified computing frame and system for industrial personalized customized production. Based on a deep neural network and reinforcement learning, a most suitable optimization algorithm is selected by automatic decision-making for a global customized production task with an industrial big data module at the bottom as an information basis, and a global optimal static scheduling plane is generated; a current dynamic event in a factory are monitored in real time; if no dynamic event requiring dynamic scheduling optimization is monitored, the global optimal static plan is executed sequentially; when a dynamic event impact requiring dynamic scheduling optimization is monitored, information of the current dynamic event is interpreted and classified, and corresponding optimization algorithms are automatically selected for dynamic scheduling optimization according to different types of dynamic events; and a dynamic scheduling scheme is evaluated by a subsequent module, an optimization scheme is regenerated or a most suitable optimization algorithm is automatically decided based on the scheme according to an evaluation result, and an equipment deployment sequence is generated for an automatic deployment. Considering the features of complicated procedures, a large amount of customization information and the high frequency of diversified dynamic events in personalized customized production, the present invention provides the adaptive-learning intelligent scheduling unified computing frame and system for industrial personalized customized production, that adopt two steps in the three aspects of static scheduling planning, dynamic scheduling planning and equipment deployment based on deep learning, that is, targeted optimization is performed after classification, thus improving the optimization efficiency and effect; and the system better fits the features of personalized customized production, and can effectively improve the efficiency of personalized customized production and minimize manual decision-making costs.

A method of localizing a host member through sensor fusion includes capturing an input image with one or more optical sensors disposed on the host member and determining a location of the host member through a global positioning system (GPS) input. The method tracks movement of the host member through an inertial measurement unit (IMU) input, generates coordinates for the host member from the GPS input and the IMU input. The method compares the input image and a high definition (HD) map input to verify distances from the host member to predetermined objects within the input image and within the HD map input. The method continuously localizes the host member by fusing the GPS input, the IMU input, the input image, and the HD map input.

A system and method of lane detection using density based spatial clustering of applications with noise (DBSCAN) includes capturing an input image with one or more optical sensors disposed on a motor vehicle. The method further includes passing the input image through a heterogeneous convolutional neural network (HCNN). The HCNN generates an HCNN output. The method further includes processing the HCNN output with DBSCAN to selectively classify outlier data points and clustered data points in the HCNN output. The method further includes generating a DBSCAN output selectively defining the clustered data points as predicted lane lines within the input image. The method further includes marking the input image by overlaying the predicted lane lines on the input image.

Disclosed herein are apparatuses, systems, methods, and computer-readable media relating to area selection in charged particle microscope (CPM) imaging. For example, in some embodiments, a CPM support apparatus may include: first logic to generate a first data set associated with an area of a specimen by processing data from a first imaging round of the area by a CPM; second logic to generate predicted parameters of the area; and third logic to determine whether a second imaging round of the area is to be performed by the CPM based on the predicted parameters of the area; wherein the first logic is to, in response to a determination by the third logic that a second imaging round of the area is to be performed, generate a second data set, including measured parameters, associated with the area by processing data from a second imaging round of the area by the CPM.

Technology is disclosed for controlling the processing of electronic communications on computing devices. An electronic communication is processed to determine mentions in the message body indicating recipients of the communication. A point-of-view (POV) is determined for each mention, with respect to recipient(s), as second-person or third-person POV. The communication is parsed into sections, and the mentions that are associated with each section are determined. Based on the POV for a mention associated with a section, it is determined that the section is directed to the recipient(s) indicated by the mention, if the POV of the mention is second-person POV, or the section is relevant to the recipient(s) indicated by the mention, if the POV of the mention is third-person POV. Additionally, an enhanced communication data is generated indicating the sections and corresponding POVs of mentions associated with the sections, and used to provide a personalized computing experience to users.

An embodiment includes training a first convolutional neural network (CNN) using a plurality of training images to generate first and second trained CNNs, and then adding an interface layer to the second trained CNN. The embodiment processes a first and second images in a sequence of images using the first trained CNN to generate a first and second result vectors. The embodiment also processes the second image using the second trained CNN and sensor data input to the interface layer to generate a third result vector. The embodiment modifies the sensor data using a compensation value. The embodiment compares the third result vector to the second result vector to generate an error value, and then calculates a modified compensation value using the error value. The embodiment then generates a sensor-compensated trained CNN based on the second trained CNN with the modified compensation value.

Embodiments described herein provide for recommending radiotherapy treatment attributes. A machine learning model predicts the preference of a medical professional and provides relevant suggestions (or recommendations) of radiotherapy treatment attributes for various categories of radiotherapy treatment. Specifically, the machine learning model predicts field geometry attributes from various field geometry attribute options for various field geometry attribute categories. The machine learning model is conditioned on patient data such as medical images and patient information. The machine learning model is trained in response to cumulative reward information associated with a medical professional accepting the provided/displayed recommendations.

Aspects of the present disclosure are directed to an artificial reality (XR) application system controlling applications in an artificial reality environment. In various cases, these controls include automatically suggesting XR applications by determining an XR context and identifying applications that match the XR context. These applications can be suggested to a user, who can authorize their execution, setting permissions for the application. In some cases, applications can be divided into components which can be progressively downloaded. By providing application suggestions relevant to the current context and progressively downloading application components, applications can appear ambient, rather than relying on users to constantly download, install, or activate applications. Permissions for applications may be revoked permanently or for certain situations—either through user permissions selections or automatically in response to determined user intents. When multiple applications are simultaneously authorized to execute, the XR application system can employ a ranking system to prevent overcrowding.