A computer system obtains, using a camera, a first plurality of images of a canopy an agricultural plot. For each respective fruit of a plurality of fruit growing in the agricultural plot, the computer system identifies a first respective image in the first plurality of images that comprises the respective fruit. The first respective image has a corresponding first camera location. The computer system also identifies a second respective image in the first plurality of images that comprises the respective fruit. The second respective image has a corresponding second camera location. The computer system uses at least i) the first and second respective images and ii) a distance between the first and second camera locations to determine a corresponding size of the respective fruit.

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for storing facial recognition image data in a cache. One of the methods includes receiving an image from a camera, detecting, in the received image, a face of a person, searching a biometric data cache based on the detected face, in response to searching the biometric data cache based on the detected face, determining whether the biometric data cache includes data for the person, in response to a determination that the biometric data cache includes data for the person, using the data from the biometric data cache to determine an identifier for the person, and in response to a determination that the biometric data cache does not include data for the person: searching a data storage system based on the detected face of the person to determine whether the data storage system includes data for the person.

Low-dimensional feature values with which semantic factors of content are ascertained are generated from relevance between sets of two types of content. Based on a relation indicator indicating a pair of groups indicating which groups are related to first types of content groups among second types of content groups, an initial feature value extracting unit 11 extracts initial feature values of the first type of content and the second type of content. A content pair selecting unit 12 selects a content pair by selecting one first type of content and one second type of content from each pair of groups indicated by the relation indicator. A feature value conversion function generating unit 13 generates feature conversion functions 31 of converting the initial feature values into low-dimensional feature values based on the content pair selected from each pair of groups.

An abnormality detection unit detects an abnormal object in target images repeatedly acquired. An abnormality type selection unit selects, for each of the target images, an abnormality type of the abnormal object from a plurality of specific abnormality types based on values of at least two basic feature amounts. A feature amount monitoring unit monitors the values of the basic feature amounts and a value of an auxiliary feature amount corresponding to the abnormality type currently selected by the abnormality type selection unit. An adjustment processing unit executes an adjustment process corresponding to the auxiliary feature amount being monitored by the feature amount monitoring unit. The abnormality type selection unit changes the abnormality type to be selected, in accordance with the change in the values of the basic feature amounts mentioned above.

An abnormality detection unit detects an abnormal object in target images repeatedly acquired. An abnormality type selection unit selects, for each of the target images, an abnormality type of the abnormal object from a plurality of specific abnormality types based on values of at least two basic feature amounts. A feature amount monitoring unit monitors the values of the basic feature amounts and a value of an auxiliary feature amount corresponding to the abnormality type currently selected by the abnormality type selection unit. An adjustment processing unit executes an adjustment process corresponding to the auxiliary feature amount being monitored by the feature amount monitoring unit. The abnormality type selection unit changes the abnormality type to be selected, in accordance with the change in the values of the basic feature amounts mentioned above.

The present invention relates to dimensionality reduction for machine learning (ML) models. Herein are techniques that individually rank features and combine features based on their rank to achieve an optimal combination of features that may accelerate training and/or inferencing, prevent overfitting, and/or provide insights into somewhat mysterious datasets. In an embodiment, a computer ranks features of datasets of a training corpus. For each dataset and for each landmark percentage, a target ML model is configured to receive only a highest ranking landmark percentage of features, and a landmark accuracy achieved by training the ML model with the dataset is measured. Based on the landmark accuracies and meta-features values of the dataset, a respective training tuple is generated for each dataset. Based on all of the training tuples, a regressor is trained to predict an optimal amount of features for training the target ML model.

Disclosed are a pose determination method and device, and a non-transitory computer storage medium. In the method, plane features of keypoints and depth features of the keypoints are extracted from a first image. Plane coordinates of the keypoints are determined based on the plane features of the keypoints. Depth coordinates of the keypoints are determined based on the depth features of the keypoints. A pose of a region of interest corresponding to the keypoints is determined based on the plane coordinates of the keypoints and the depth coordinates of the keypoints.

In an image recognition method, a terminal determines, based on first positioning information, target object information corresponding to building information in a to-be-recognized image in desensitized map data. The desensitized map data does not include a sensitive building. Then, when the terminal determines that the target object information does not include the building information, the terminal determines that the map data does not include the building information. In this case, the terminal determines to recognize the building information as a sensitive building. In other words, the terminal may recognize, by using the desensitized map data, building information corresponding to a sensitive building in the to-be-recognized image.

In an image recognition method, a terminal determines, based on first positioning information, target object information corresponding to building information in a to-be-recognized image in desensitized map data. The desensitized map data does not include a sensitive building. Then, when the terminal determines that the target object information does not include the building information, the terminal determines that the map data does not include the building information. In this case, the terminal determines to recognize the building information as a sensitive building. In other words, the terminal may recognize, by using the desensitized map data, building information corresponding to a sensitive building in the to-be-recognized image.

A classifier network has at least two distinct sets of refined data, wherein the first two sets of refined data are sets of numbers representing the features values data received from sensors or a manufactured part. Performing, via at least two distinct types of support vector machines using an associated feature selection process for each classifier independently in a first layer, anomaly detection on the manufactured part. Then, using the stored data including refined data of at least two different types of data transforms and performing, via at least a two distinct types of support vector machines in a second layer, an associated feature selection process for each classifier independently. Forming at least four distinct compound classifier types for anomaly detection on the part using the stored data or coefficients. The ensemble of second layer support vector machine outputs compare the results to determine the presence of an anomaly.