Embodiments are directed to techniques for image content extraction. Some embodiments include extracting contextually structured data from document images, such as by automatically identifying document layout, document data, document metadata, and/or correlations therebetween in a document image, for instance. Some embodiments utilize breakpoints to enable the system to match different documents with internal variations to a common template. Several embodiments include extracting contextually structured data from table images, such as gridded and non-gridded tables. Many embodiments are directed to generating and utilizing a document template database for automatically extracting document image contents into a contextually structured format. Several embodiments are directed to automatically identifying and associating document metadata with corresponding document data in a document image to generate a machine-facilitated annotation of the document image. In some embodiments, the machine-facilitated annotation may be used to generate a template for the template database.

A system for cross-modal data retrieval is provided that includes a neural network having a time series encoder and text encoder which are jointly trained using an unsupervised training method which is based on a loss function. The loss function jointly evaluates a similarity of feature vectors of training sets of two different modalities of time series and free-form text comments and a compatibility of the time series and the free-form text comments with a word-overlap-based spectral clustering method configured to compute pseudo labels for the unsupervised training method. The computer processing system further includes a database for storing the training sets with feature vectors extracted from encodings of the training sets. The encodings are obtained by encoding a training set of the time series using the time series encoder and encoding a training set of the free-form text comments using the text encoder.

Disclosed is a method for short-term traffic risk prediction of road sections by using roadside observation data. The method includes the following steps: 1) vehicle trajectory data in the detection area is obtained by using roadside observation data; 2) according to the continuous driving trajectories in the detection area, the traffic flow indicators are counted, and the surrogate safety indicators between vehicles are calculated; 3) time to collision and deceleration are selected as identification indicators to identify conflict events with collision risk in the detection area; 4) traffic flow indicators and surrogate safety indicators within the set time before the occurrence of conflict events are extracted, and the feature screening of various extracted indicators is performed by using classification algorithms; 5) based on the selected feature indicators, the indicators with the highest importance ranking are selected as the input to build a short-term traffic risk prediction model, and the model training and testing are completed by using the identified conflict events; 6) the short-term traffic risk prediction model is used to predict the risk of road sections. The proposed method can improve the prediction accuracy rate of road sections.

An information processing apparatus includes an acquisition unit, a calculation unit, and a generation unit. The acquisition unit acquires information including information regarding multiple nodes and information regarding multiple links connecting the multiple nodes and acquires constraint information regarding node pairs included in the multiple nodes. The constraint information includes a positive constraint and a negative constraint. The calculation unit calculates, for each of multiple clusters, a classification proportion into which the multiple nodes are classified and calculates a degree of importance of each of the multiple clusters. The classification proportion represents a proportion in which each of the multiple nodes is classified as one of the multiple clusters. The generation unit generates a probability model for performing probabilistic clustering on the multiple nodes. The probability model is generated by using at least each of the information regarding the links, the constraint information, the classification proportion, and the degree of importance.

Systems for sorting seeds are disclosed, as well as batches of seeds that have been sorted using the systems.

A device for and computer implemented method of image content recognition and of training a neural network for image content recognition. The method comprising collecting a first set of digital images from a database, the first set of digital images is sampled from digital images assigned to a many shot class; creating a first training set comprising the collected first set of digital images; training a first artificial neural network comprising a first feature extractor and a first classifier for classifying digital images using the first training set; collecting first parameters of the trained first feature extractor, collecting second parameters of the trained classifier, determining third parameters of a second feature extractor of a second artificial neural network depending on the first parameters, determining fourth parameters of a second classifier for classifying digital images of the second artificial neural network.

A device for and computer implemented method of image content recognition and of training a neural network for image content recognition. The method comprising collecting a first set of digital images from a database, the first set of digital images is sampled from digital images assigned to a many shot class; creating a first training set comprising the collected first set of digital images; training a first artificial neural network comprising a first feature extractor and a first classifier for classifying digital images using the first training set; collecting first parameters of the trained first feature extractor, collecting second parameters of the trained classifier, determining third parameters of a second feature extractor of a second artificial neural network depending on the first parameters, determining fourth parameters of a second classifier for classifying digital images of the second artificial neural network.

A system and a method for supporting automated valet parking, and an infrastructure and a vehicle therefor are provided. The method for operating a vehicle which supports automated valet parking includes initializing automated valet parking and receiving, from an infrastructure, a target position which is related to the vehicle and a guide route which guides movement to the target position. Additionally, the method includes performing automated driving by the vehicle along the guide route, and measuring a position of the vehicle based on behavior information of the vehicle and environmental information, while the vehicle performs the automated driving. The environmental information includes at least one among parking lot slot information, road signs, walls, pillars, and floor markings.

Various embodiments are provided for implementing an approximation nearest neighbour (ANN) search in a computing environment are provided. An approximation nearest neighbour (ANN) of a plurality of feature vectors in hyper-planes with dynamically variable subspaces by searching an inverted index may be retrieved.

Computerized methods to predict or determine if a receptor is associated with a biological process in a target tissue are provided. Methods of training and using a machine learning model to predict or determine if a receptor is associated with a biological process in a target tissue as well as systems and computer program product to do same are also provided.