Abstract: Uncertainty-based data filtering in a vehicle may include acquiring sensor data from a plurality of sensors of the autonomous vehicle; applying a first machine learning model to the sensor data; determining that a result of the first machine learning model disagrees with another result; and transmitting, based on the result of the first machine learning model disagreeing with the other result, the sensor data to a server.

Abstract: Features are extracted and tabulated for transactions on a per-operator basis. Metrics are derived from the features on a per-feature basis and on an aggregate bases for the features as a whole. Additional metrics are calculated on a per-feature basis and an aggregate bases for a group of operators. A current transaction being conducted in real time by a given operator has current features extracted and a risk score is generated. The risk score is compared against the metrics associated with the given operator and the group of operators. A determination is made as to whether the current transaction is fraud or is not fraud based on the comparison. When fraud is detected, an alert is raised, and the current transaction provided to a fraud system for further fraud evaluation.

Abstract: A feedback processing module includes a memory configured to store feedback received from an encoder. The feedback includes parameters associated with encoded graphics content generated by a graphics engine. The feedback processing module also includes a processor configured to generate configuration information for the graphics engine based on the feedback. The graphics engine is configured to execute a workload based on the configuration information. In some cases, the feedback processing module is also configured to receive feedback from a decoder that is used to decode the graphics content that is encoded by the encoder and generate the configuration information based on the feedback received from the decoder.

Abstract: The disclosed embodiments describe methods, systems, and apparatuses for determining user intent. A method is disclosed comprising obtaining a session text of a user; calculating, by the processor, a feature vector based on the session text; determining probabilities that the session text belongs to a plurality of intent labels, the probabilities calculated using a multi-level hierarchal intent classification model, the intent labels assigned to levels in the multi-level hierarchal intent classification model; and assigning a user intent to the session text based on the probabilities.

Abstract: A method for image segmentation includes receiving an input image. The method further includes obtaining a deep learning model having a triad of predictors. Furthermore, the method includes processing the input image by a shape model in the triad of predictors to generate a segmented shape image. Moreover, the method includes presenting the segmented shape image via a display unit.

Abstract: Disclosed are a system and method for quantifying changes in animal posture from video recordings of multiple interacting animals. Also disclosed are a system and method for training a posture identification algorithm efficiently with minimal human effort and evaluating the posture of millions of animal images. The disclosed process employs video recordings of behaving animals to produce a full set of geometrical coordinates for each body part over time. Inter alia, the disclosed system should be useful for organizations interested in doing large small-molecule screens to find drugs that affect the behavior of animals, and for those investigating new behavioral diagnostics for human patients. The disclosed system allows one to record video of animals and then automatically calculate the position of each body part over time. The output of this calculation can be used for simple analysis of the amount of movement for specific body parts, or part of a more complex scheme for measuring behaviors.

Abstract: A multispectral sensor device may include a sensor array comprising a plurality of channels and one or more processors to determine that a time-sensitive measurement is to be performed, wherein the time-sensitive measurement is to be performed using data collected by one or more channels of the plurality of channels; cause the data to be collected by a proper subset of channels, of the plurality of channels, wherein the proper subset of channels includes the one or more channels; and determine the time-sensitive measurement based on the data.

Abstract: Apparatus, methods, and computer-readable media for providing frameworks for data source representation and compression using an index file format are disclosed herein. The index file format separate information about symbols in a data file and information about the corresponding location of those symbols in the data file. The described techniques provide mechanisms for reducing the size associated with the representation of the symbols information and/or the size associated with the representation of the location information.

Abstract: Obtain, from an existing machine learning classifier, original probabilistic scores classifying samples taken from two or more groups into two or more classes via supervised machine learning. Associate the original probabilistic scores with a plurality of original Lagrange multipliers. Adjust values of the plurality of original Lagrange multipliers via low-dimensional convex optimization to obtain updated Lagrange multipliers that satisfy fairness constraints as compared to the original Lagrange multipliers. Based on the updated Lagrange multipliers, closed-form transform the original probabilistic scores into transformed probabilistic scores that satisfy the fairness constraints while minimizing loss in utility. The fairness constraints are with respect to the two or more groups.

Abstract: Method and apparatus for mechanical sex-sorting of mosquitoes by extracting a class of mosquitoes from unsorted mosquitoes comprises obtaining unsorted mosquitoes, obtaining images of individual mosquitoes in a stationary phase, electronically classifying the individuals from the images into male mosquitoes and/or female mosquitoes, and possibly also unclassified objects; obtaining co-ordinates of individuals of at least one of the male mosquito and female mosquito classifications, and using a robot arm to reach an individual identified by the obtained coordinates to store or remove the individuals, thereby to provide sex-sorted mosquitoes.

Abstract: The present description relates to systems, methods, and non-transitory computer readable media for generating digital responses to digital queries by utilizing a classification model and query-specific analysis models. For example, the described systems can train a classification model to generate query classifications corresponding to product queries, conversational queries, and/or recommendation/purchase queries. Moreover, the described systems can apply the classification model to select pertinent models for particular queries. For example, upon classifying a product query, the described systems can utilize a neural ranking model (trained based on a set of training product specifications and training queries) to generate relevance scores for product specifications associated with a digital query.

Abstract: A metadata extraction machine accesses an image that depicts an item. The item depicted in the image may have an attribute that describes a characteristic of the item and an attribute descriptor that corresponds to the attribute of the item and specifies a value of the attribute. The metadata extraction machine performs an analysis of the image. The analysis may include identifying the attribute descriptor corresponding to the attribute based on image segmentation of the image. The metadata extraction machine transmits a communication to a device of a user based on the identifying of the attribute descriptor corresponding to the attribute of the item depicted in the image.

Abstract: A diagnostic tool for deep learning similarity models and image classifiers provides valuable insight into neural network decision-making. A disclosed solution generates a saliency map by: receiving a baseline image and a test image; determining, with a convolutional neural network (CNN), a first similarity between the baseline image and the test image; based on at least determining the first similarity, determining, for the test image, a first activation map for at least one CNN layer; based on at least determining the first similarity, determining, for the test image, a first gradient map for the at least one CNN layer; and generating a first saliency map as an element-wise function of the first activation map and the first gradient map. Some examples further determine a region of interest (ROI) in the first saliency map, cropping the test image to an area corresponding to the ROI, and determine a refined similarity score.

Abstract: Systems and methods are disclosed for model based document image enhancement. Instead of requiring paired dirty and clean images for training a model to clean document images (which may cause privacy concerns), two models are trained on the unpaired images such that only the dirty images are accessed or only the clean images are accessed at one time. One model is a first implicit model to translate the dirty images from a source space to a latent space, and the other model is a second implicit model to translate the images from the latent space to clean images in a target space. The second implicit model is trained based on translating electronic document images in the target space to the latent space. In some implementations, the implicit models are diffusion models, such as denoising diffusion implicit models based on solving ordinary differential equations.

Abstract: This disclosure relates generally to categorical time-series clustering. In an embodiment, the method for categorical time-series clustering for categorical time-series associated with distinct subjects obtained from sensors. Based on the categorical time-series, the subjects are clustered into clusters by using a Markov chain model. Clustering the subjects include assigning each subject to a cluster. The subjects are assigned to the clusters by determining cluster-specific transition matrices based on a transitional probability of the subject's transitioning between states. A semi-distance function is constructed for each cluster-specific transitional matrix between the states at multiple time instances, which us indicative of a conditional probability of movement of the subject between the states at different time instance.

Abstract: A product defect detection method, device and system are disclosed.

Abstract: This invention relates to a method of blocking or passing messages sent via a firewall, based on parsing, of symbols strings contained in said messages, among different keywords, assigning to said messages either a blocking class or a passing class, comprising: performing a dimensionality reduction step for said different classes on a training set of messages whose classes are known, and then classifying one or more unknown messages among said different classes with reduced dimensionality, said dimensionality reduction step being performed on said training set of messages by machine learning including processing, for several first matrices and for several second matrices, a parameter representative of a product of two first and second matrices to assess to which given class a given message belongs: first matrix representing a first array of keywords versus symbols strings contained in a first given message, second matrix representing the values of differences between said first array and a second array of key

Abstract: Disclosed herein are system, method, and computer program product embodiments for distinguishing a disease state from a non-disease state in an image. An embodiment operates by receiving an image of a target area over a network. The embodiment then corrects for background noise in the image by applying a semantic segmentation filter to obtain a segmented image. The sematic segmentation filter may be trained to remove the background noise from the image. The embodiment then determines, using a trained artificial intelligence (AI) model and the segmented image, at least one classification for the target area. The embodiment finally causes the display of the at least one classification and disease information on a user device associated with a user. The trained AI model may be trained using at least augmented images obtained from a set of images to correct for at least an imbalance in the set of images.

Abstract: Computer software development has produced many advances within computer science and in most aspects of modern society. Even with modern quality control, bug finding, and other code checking applications, computer software is often less than ideal. A developer may write code that is functionally accurate but lacks security, documentation, speed, storage, reusability, or other element that may make a segment of software code less than ideal. Identifying equivalent code, within a defined hypothesis strength and/or resource limitation, and, when found, replacing it with a vetted equivalent promotes the deployment of software that is more robust, secure, usable and reusable, and/or satisfies performance or other objectives.

Abstract: A method of improving image quality of a stream of input images is described. The stream of input images, including a current input image, is received. One or more target objects, including a first target object, are identified spatio-temporally within the stream of input images. The one or more target objects are tracked spatio-temporally within the stream of input images. The current input image is segmented into i) a foreground including the first target object, and ii) a background. The foreground is processed to have improved image quality in the current input image. Processing of the foreground further comprises processing the first target object using a same processing technique as for a prior input image of the stream of input images based on the tracking of the first target object. The background is processed differently from the foreground. An output image is generated by merging the foreground with the background.

Abstract: Techniques for determining a probability of a false negative associated with a location of an environment are discussed herein. Data from a sensor, such as a radar sensor, can be received that includes point cloud data, which includes first and second data points. The first data point has a first attribute and the second data point has a second attribute. A difference between the first and second attributes is determined such that a frequency distribution may be determined. The frequency distribution may then be used to determine a distribution function, which allows for the determination of a resolution function that is associated with the sensor. The resolution function may then be used to determine a probability of a false negative at a location in an environment. The probability can be used to control a vehicle in a safe and reliable manner.

Abstract: Provided are a processing method, a terminal device, and a medium. The method includes the steps described below. Basic information of a target video is determined; attribute information of the target video is determined based on the basic information; in a case where the attribute information indicates that the target video is a video capable of being structured, chapter division is performed on the target video based on the basic information to obtain at least two video clips; and chapter description information of the at least two video clips, a key frame of the at least two video clips and video description information of the target video are determined.

Abstract: A method of reducing false positives and identifying relevant true alerts in a video management system includes analyzing images to look for patterns indicating changes between subsequent images. When a pattern indicating changes between subsequent images is found, the video management system solicits from a user an indication of whether the pattern belongs to one of two or more predefined categories. The patterns indicating changes between subsequent images are saved for subsequent use. Subsequent images received from the video camera are analyzed to look for patterns indicating changes between subsequent images. When a pattern indicating changes between subsequent images is detected by the video management system, the video management system compares the pattern indicating changes between subsequent images to those previously categorized into one of the two or more predefined categories. Based on the comparison, the video management system may provide an alert to the user.

Abstract: Systems and methods for vehicle-to-vehicle communications are provided. An example computer-implemented method includes obtaining from a first autonomous vehicle, by a second autonomous vehicle, a first compressed intermediate environmental representation. The first compressed intermediate environmental representation is indicative of at least a portion of an environment of the second autonomous vehicle. The method includes generating a first decompressed intermediate environmental representation by decompressing the first compressed intermediate environmental representation. The method includes determining, using one or more machine-learned models, an updated intermediate environmental representation based at least in part on the first decompressed intermediate environmental representation and a second intermediate environmental representation generated by the second autonomous vehicle.

Abstract: Systems and methods for trajectory prediction and motion device control include receiving a first set of data about a surrounding environment of a motion device. The first set of data includes first sensory data and first position data. The method includes receiving a second set of data about the surrounding environment of the motion device. The second set of data includes second sensory data and second position data, where the first sensory data differs in modality from the second sensory data. The method also includes extracting features from the first set of data that model motion behavior of the motion device, and extracting features from the second set of data that model motion behavior of the motion device. The method includes inputting the features into a machine learning model, and computing a first trajectory based on a first latent variable and a second trajectory based on a second latent variable.

Abstract: Methods, apparatuses, and non-transitory machine-readable media for displaying information and/or images on a display of a computing device based on received data. Apparatuses can include a display screen, a memory resource, a recognition sensor, and a controller. An example controller can receive data and activate information and/or images on a display screen based in part on the received data. In another example, a method can include storing recognition data in a memory resource, receiving primary recognition data, comparing the primary recognition data to the stored recognition data, and activating the display screen for a viewing angle responsive to authentication of the primary recognition data through the comparison of the primary recognition data and the stored recognition data.

Abstract: A dual sensor imaging system and a privacy protection imaging method thereof are provided. The system is configured to control at least one color sensor and at least one IR sensor to respectively capture multiple color images and multiple IR images by adopting multiple exposure conditions adapted for an imaging scene, adaptively select a combination of the color image and the IR image that can reveal details of the imaging scene, detect a feature area with features of a target of interest in the color image, and fuse the color image and the IR image to generate a fusion image with the details of the imaging scene, and crop an image of the feature area of the fusion image to be replaced with an image not belonging to the IR image, so as to generate a scene image.

Abstract: Provided are an AI system for simulating functions such as recognition, determination, and so forth of human brains by using a mechanical learning algorithm like deep learning, or the like, and an application thereof. In particular, according to the AI system and the application thereof, a neural network training method includes obtaining a plurality of first images belonging to a particular category and a plurality of second images for which a category is not specified, training a neural network model for category recognition, based on the plurality of first images belonging to the particular category, recognizing at least one second image corresponding to the particular category among the plurality of second images, by using the trained neural network model, and modifying and refining the trained neural network model based on the recognized at least one second image.

Abstract: A method, a computer system, and a computer program product are provided for federated learning. An aggregator may receive cluster information from distributed computing devices. The cluster information may relate to identified clusters in sample data of the distributed computing devices. The cluster information may include centroid information per cluster. The aggregator may include a processor. The aggregator may integrate the cluster information to define data classes for machine learning classification. The integrating may include computing a respective distance between centroids of the clusters in order to determine a total number of the data classes. The aggregator may send a deep learning model that includes an output layer that has a total number of nodes equal to the total number of the data classes. The deep learning model may be for the distributed computing devices to perform machine learning classification in federated learning.

Abstract: An object tracking method and apparatus are provided. The object tracking method includes detecting a target object in a first-type input image that is based on light in a first wavelength band, tracking the target object in the first-type input image based on detection information of the target object, measuring a reliability of the first-type input image by comparing the first-type image to an image in a database, comparing the reliability of the first-type input image to a threshold, and tracking the target object in a second-type input image that is based on light in a second wavelength band.

Abstract: A system for viewing at a client device a series of three-dimensional virtual views over the Internet of a volume visualization dataset contained on centralized databases employs a transmitter for securely sending volume visualization dataset from a remote location to the centralized database, more than one central data storage medium containing the volume visualization dataset, and a plurality of servers in communication with the centralized databases to create virtual views based on client requests. A resource manager load balances the servers, a security device controls communications between the client device and server and the resource manager and central storage medium. Physically secured sites house the components. A web application accepts at the remote location user requests for a virtual view of the volume visualization dataset, transmits the request to the servers, receives the resulting virtual view from the servers, and displays the resulting virtual view to the remote user.

Abstract: A computer-implemented method for automatically detecting events and creating photo-product designs based on the events in a photo-product design system includes automatically identifying an event by an event detection module based on daily numbers of captured photos over a plurality of days, automatically selecting a photo-product type by an intelligent product design creation engine in the photo-product design system, calculating a daily weight for a photo product design in the photo-product type based on the daily numbers of captured photos, automatically determining a number of product photos allocated to each day based on associated daily weight, automatically selecting product photos from the captured photos each day at the event according to the number of product photos allocated to each day, and automatically creating a photo-product design for the event using the selected product photos.

Abstract: An electronic device for active learning includes at least one memory and at least one processor coupled to the at least one memory. The at least one processor is configured to select one or more entries from a data set including unlabeled data based on a similarity between the one or more entries and labeled data. The at least one processor is further configured to cause the one or more entries to be labeled.

Abstract: The present invention pertains to a method for automatic sea lice monitoring in fish aquaculture, the method comprising submerging a camera (4) in a sea pen (300) comprising fish, using the camera to make an image of at least one of said fish, analysing the image to differentiate between individual sea lice present on the fish and the fish itself and assessing the number of sea lice present on the fish, wherein the camera is attached to a device (1, 10, 100) for guiding the salmon along an imaging track (5), the camera being directed to the track.

Abstract: An information processing apparatus (100) includes an input unit (102) that inputs a first threshold value and a second threshold value which are threshold values related to a degree of similarity of a feature value of each of a plurality of pieces of data, the first threshold value being for regarding data as belonging to an identical group and the second threshold value being smaller than the first threshold value, and a grouping unit (104) that groups the data by using the degree of similarity, the first threshold value, and the second threshold value, in which the grouping unit (104) causes data of which the degree of similarity with representative data is greater than the first threshold value to be included in the same group, and selects new representative data from among pieces of data of which the degree of similarity with the representative data already present is less than the second threshold value.

Abstract: Techniques for distinguishing objects (e.g., an individual or an individual pushing a shopping cart) are disclosed. An object is detected in images of a scene. A height map is generated from the images, and the object is represented as height values in the height map. Based on height properties associated with another object, it is determined whether the other object is associated with the object. If so determined, the objects are classified separately.

Abstract: A device, apparatus, and method provide logic for processing information. In one implementation, a device may include an image acquisition unit configured to acquire an image, and a transmission unit configured to transmit information associated with the image to an information processing apparatus, such as a server. The server may be associated with a first feature quantity dictionary. The device also may include a receiving unit configured to receive a second feature quantity dictionary from the server in response to the transmission. The second feature quantity dictionary may include less information than the first feature quantity dictionary, and the server may generate the second feature quantity dictionary based on the image information and the first feature quantity dictionary. The device may include an identification unit configured to identify an object within the image using the second feature quantity dictionary.

Abstract: The present disclosure provides an improved approach to implement structure learning of neural networks by exploiting correlations in the data/problem the networks aim to solve. A greedy approach is described that finds bottlenecks of information gain from the bottom convolutional layers all the way to the fully connected layers. Rather than simply making the architecture deeper, additional computation and capacitance is only added where it is required.

Abstract: A processing node in a temporal memory system includes a spatial pooler and a sequence processor. The spatial pooler generates a spatial pooler signal representing similarity between received spatial patterns in an input signal and stored co-occurrence patterns. The spatial pooler signal is represented by a combination of elements that are active or inactive. Each co-occurrence pattern is mapped to different subsets of elements of an input signal. The spatial pooler signal is fed to a sequence processor receiving and processed to learn, recognize and predict temporal sequences in the input signal. The sequence processor includes one or more columns, each column including one or more cells. A subset of columns may be selected by the spatial pooler signal, causing one or more cells in these columns to activate.

Abstract: Automated detection of reasoning in arguments. A training set is generated by: obtaining multiple arguments, each comprising one or more sentences provided as digital text; automatically estimating a probability that each of the arguments includes reasoning, wherein the estimating comprises applying a contextual language model to each of the arguments; automatically labeling as positive examples those of the arguments which have a relatively high probability to include reasoning; and automatically labeling as negative examples those of the arguments which have a relatively low probability to include reasoning. Based on the generated training set, a machine learning classifier is automatically trained to estimate a probability that a new argument includes reasoning. The trained machine learning classifier is applied to the new argument, to estimate a probability that the new argument includes reasoning.

Abstract: There is described a computer-implemented method for generating a vector representation of an image, the computer-implemented method comprising: receiving a given image and semantic information about the given image; generating a first vector representation of the given image using an image embedding method; generating a second vector representation of the semantic information using a word embedding method; combining the first vector representation of the image to be embedded and the second vector representation of the semantic information together, thereby obtaining a modified vector representation for the image to be embedded; and outputting the modified vector representation.

Abstract: A processing system for an unmanned vehicle (UV) such as an unmanned aerial vehicle (UAV) is provided. The processing system comprises a first processing unit of an integrated circuit and a second processing unit of the integrated circuit. The processing system comprises a first operating system provisioned using the first processing unit. The first operating system is configured to execute a first vehicle control process. The processing system comprises a virtualization layer configured using at least the second processing unit, and a second operating system provisioned using the virtualization layer. The second operating system is configured to execute a second vehicle control process.

Abstract: An electronic device according to various embodiments, may include a camera, a communication circuit, a display and a processor, and the processor may be configured to acquire an image of an external object, transmit the image to an external electronic device, so that the external electronic device transmits product information related to the external object included in the image to the electronic device, using the communication circuit, determine one or more keywords related to the product information transmitted from the external electronic device, identify preference category information and preference brand information of a user, based on a usage record related to a product of the user related to the electronic device, determine weights for the one or more keywords, using the preference category information and the preference brand information, and provide at least some keywords having weights which satisfy a specified condition among the one or more keywords, on the display together with the product info

Abstract: Methods and devices for manipulating an image are described. The method comprises receiving image data, the image data including a first image obtained from a first camera and a second image obtained from a second camera, the first camera and the second camera being oriented in a common direction; identifying one or more boundaries of an object in the image data by analyzing the first image and the second image; and displaying a manipulated image based on the image data, wherein the manipulated image includes manipulation of at least a portion of the first image based on boundaries of the object.

Abstract: A system and method for inserting a composited image or otherwise generated graphic into a selected video by way of a programmatic process. According to some embodiments, a system may comprise an Automated Placement Opportunity Identification (APOI) engine, a Placement Insertion Interface (PII) engine, a preview system, and an automated compositing service. The system finalizes a graphic composite into a video and provides a user with a preview for final export or further manipulation.

Abstract: Various disclosed embodiments are directed to classify or determining an image style of a target image according to a consumer application based on determining a similarity score between the image style of a target image and one or more other predetermined image styles of the consumer application. Various disclosed embodiments can resolve image style transfer destructiveness functionality by making various layers of predetermined image styles modifiable. Further various embodiments resolve tedious manual user input requirements and reduce computing resource consumption, among other things.

Abstract: A computer-implemented method for annotating images is disclosed. The computer-implemented method includes generating a saliency map corresponding to an input image, wherein the input image is an image that requires annotation, generating a behavior saliency map, wherein the behavior saliency map is a saliency map formed from an average of a plurality of objects contained within respective bounding boxes of a plurality of sample images, generating a historical saliency map, wherein the historical saliency map is a saliency map formed from an average of a plurality of tagged objects in the plurality of sample images, fusing the saliency map corresponding to the input image, the behavior saliency map, and the historical saliency map to form a fused saliency map, and generating, based on the fused saliency map, a bounding box around an object in the input image.

Abstract: Methods, systems, and computer programs are presented for adding new features to a network service. A method includes receiving an image depicting an object of interest. A category set is determined for the object of interest and an image signature is generated for the image. Using the category set and the image signature, the method identifies a set of publications within a publication database and assigns a rank to each publication. The method causes presentation of the ranked list of publications at a computing device from which the image was received.

Abstract: Systems and methods are disclosed for applying neural networks in resource-constrained environments. A system may include a sensor located in a resource-constrained environment configured to generate sensor data of the resource-constrained environment. The system may also include a first computing device not located in the resource-constrained environment configured to produce a neural network structure based on the sensor data. The system may further include a second computing device located in the resource-constrained environment configured to provide the sensor data as input to the neural network structure. The second computing device may be further configured to determine a state of the resource-constrained environment based on the input of the sensor data to the neural network structure.

Abstract: In one embodiment, the present disclosure is directed to a method of verifying a location of an organization. The method includes providing a user interface to a device of a representative of an organization, the user interface prompting the representative to provide organization data. The organization data includes organization name data indicative of an organization name, organization address data indicative of a location of the organization, photograph files for photographs of specific different portions of the location of the organization, and photograph location data indicative of where each of the photographs was taken. The method further includes receiving the organization data from the device, and verifying that the organization address data corresponds with the organization address data.