Abstract: This application provides a voice interaction processing method and apparatus, to achieve a friendly and natural voice interaction effect and reduce power consumption. In the method, a microprocessor enables an image collector only when determining, based on voice data collected by a voice collector, that a first user is a target user; then the image collector collects user image data and transmits the user image data to the microprocessor; and the microprocessor sends a wakeup instruction to an application processor only when determining, based on the user image data, that the target user is in a voice interaction state. Based on the foregoing method, nus-enabling of the image collector and the application processor is avoided to some extent, and power consumption is reduced.

Abstract: Various examples of the present disclosure include a stereoscopic deep neural network (DNN) that produces accurate and reliable results in real-time. Both LIDAR data (supervised training) and photometric error (unsupervised training) may be used to train the DNN in a semi-supervised manner. The stereoscopic DNN may use an exponential linear unit (ELU) activation function to increase processing speeds, as well as a machine learned argmax function that may include a plurality of convolutional layers having trainable parameters to account for context. The stereoscopic DNN may further include layers having an encoder/decoder architecture, where the encoder portion of the layers may include a combination of three-dimensional convolutional layers followed by two-dimensional convolutional layers.

Abstract: Disclosed are a method and a system for determining a response to a digital task in a computer-implemented crowd-sourced environment. The method comprises determining if a number of the plurality of responses to the digital task received meets a pre-determined minimum answer threshold; in response to the number of the plurality of responses to the digital task meeting the pre-determined minimum answer threshold, executing: for each of the plurality of responses generating, by the server, a confidence parameter representing a probability of an associated one of the plurality of responses being correct; ranking the plurality of responses based on the confidence parameter to determine a top response being associated with a highest confidence parameter; and in response to the highest confidence parameter being above a pre-determined minimum confidence threshold, assigning a value of the top response as a label for the digital task and terminating the digital task execution.

Abstract: In a prediction model generating apparatus 1, data obtaining unit 11 obtains text data; variable group classifying unit 12 classifies the data into a plurality of variable groups; variable scoring unit 13 scores the data of at least one of the plurality of variable groups by associating that data with the data of another group; variable input unit 14 takes the data of the scored group as a response variable, and the data of the other group associated with the scored group as an explaining variable, and inputs those data to machine learning unit 15. The machine learning unit 15 generates, through machine learning, a prediction model predicting the response variable from the explaining variable.

Abstract: A process for recognizing and responding to a detected audio inquiry by gathering information distributed amongst users and providing a calculated result includes monitoring audio communications and detecting an audio inquiry from a first electronic communication device. It is determined then the audio inquiry is of a distributed information gathering and mathematical operation type and, in response, a plurality of target electronic communication devices that are capable of providing partial answers to the audio inquiry are identified and a request for a partial answer transmitted to each. Corresponding partial answers are received and a mathematical operation performed on the partial answers to generate a final answer to the audio inquiry different from the corresponding partial answers. The final answer is then provided to the first electronic communication device.

Abstract: Techniques are described for a contextual natural language understanding (cNLU) framework that is able to incorporate contextual signals of variable history length to perform joint intent classification (IC) and slot labeling (SL) tasks. A user utterance provided by a user within a multi-turn chat dialog between the user and a conversational agent is received. The user utterance and contextual information associated with one or more previous turns of the multi-turn chat dialog is provided to a machine learning (ML) model. An intent classification and one or more slot labels for the user utterance are then obtained from the ML model. The cNLU framework described herein thus uses, in addition to a current utterance itself, various contextual signals as input to a model to generate IC and SL predictions for each utterance of a multi-turn chat dialog.

Abstract: An image forming apparatus includes a communication unit configured to communicate with an external controller, a reception unit configured to receive, in a case where a predetermined error related to the communication occurs, log information on the predetermined error from the external controller, and a storage control unit configured to store, in a storage unit of the image forming apparatus, the log information received from the external controller and log information indicating an operation of the image forming apparatus corresponding to a timing at which the predetermined error has occurred.

Abstract: A method for determining at least one artifact calibration coefficient is provided. The method may include obtaining preliminary projection values of a first object. The radiation rays may be detected by at least one radiation detector. The method may further include generating a preliminary image of the first object based on the preliminary projection values of the first object and generating calibrated projection values of the first object based on the preliminary image. The method may further include determining a relationship between the preliminary projection values and the calibrated projection values. The method may further include, for each of the at least one radiation detector, determining a location of the radiation detector and determining an artifact calibration coefficient corresponding to the radiation detector based on the relationship between the preliminary projection values and the calibrated projection values and the location of the radiation detector.

Abstract: Various embodiments of the present invention pertain to an electronic device and an operation method therefor.

Abstract: A method for computing a human-machine hybrid ensemble prediction includes: receiving an individual forecasting question (IFP); classifying the IFP into one of a plurality of canonical question topics; identifying machine models associated with the canonical question topic; for each of the machine models: receiving, from one of a plurality of human participants: a first task input including a selection of sets of training data; a second task input including selections of portions of the selected sets of training data; and a third task input including model parameters to configure the machine model; training the machine model in accordance with the first, second, and third task inputs; and computing a machine model forecast based on the trained machine model; computing an aggregated forecast from machine model forecasts computed by the machine models; and sending an alert in response to determining that the aggregated forecast satisfies a threshold condition.

Abstract: In a Multi-Function Peripheral (MFP), identification information readable from an IC card is accepted, an administrator ID and an administrator password are used to access a policy server, a policy is registered in the policy server based on the accepted identification information, and storage of an encrypted file including image data obtained by a scan is controlled. A security setting in accordance with the policy based on the identification information is associated with the encrypted file, and a restriction of at least a part of an operation on the file is enabled in accordance with the security setting in accordance with the policy.

Abstract: In one aspect, a device may include at least one processor and storage accessible to the at least one processor. The storage may include instructions executable by the at least one processor to categorize a first image into one of a work category and a non-work category. Based on the first image being categorized into the work category, the instructions may be executable to present the first image on a display while the device is disposed at a first location associated with the work category. Based on the first image being categorized into the non-work category, the instructions may be executable to present the first image on the display while the device is disposed at a second location associated with the non-work category.

Abstract: A learning device includes one or more processors. The processors input, to an input layer of a neural network including hidden layers defined for respective first arrangement patterns indicating arrangement of one or more words, and output layers connected with some of hidden layers, one or more first morphemes conforming to any of first arrangement patterns, among morphemes included in a document, and learn the neural network to minimize a difference between one or more second morphemes conforming to any of second arrangement patterns indicating arrangement of one or more words, among morphemes included in the document, and output morphemes from the neural network for the input first morphemes. The processors output an embedding vector of the first morphemes that is obtained based on a weight of the learned neural network.

Abstract: The disclosed privacy management platforms are adapted to scan any number of data sources in order to provide users with visibility into stored personal information, risk associated with storing such information and/or usage activity relating to such information. The platforms may correlate personal information findings to specific data subjects and may employ machine learning models to classify findings as corresponding to a particular personal information attribute to provide an indexed inventory across multiple data sources.

Abstract: Embodiments of the disclosure provide systems and methods for domain adaptation between a plurality of source domains and a target domain. The artificial intelligence method includes receiving labeled data from the plurality of source domains and unlabeled data from the target domain. The method further includes separately training, by a processor, a plurality of source classifiers each corresponding to a source domain using the labeled data received from the respective source domains. The method also includes selecting a subset of the labeled data received from each source domain based on a similarity between the selected labeled data and the unlabeled data of the target domain. The method additionally includes refining, by the processor, each source classifier using the selected subset of the labeled data, and predicting labels of the unlabeled data using the refined source classifiers.

Abstract: In various example embodiments, a system and method for a Listing Engine that translates a first listing from a first language to a second language. The first listing includes an image(s) of a first item. The Listing Engine provides as input to an encoded neural network model a portion(s) of a translated first listing and a portions(s) of a second listing in the second language. The second listing includes an image(s) of a second item. The Listing Engine receives from the encoded neural network model a first feature vector for the translated first listing and a second feature vector for the second listing. The first and the second feature vectors both include at least one type of image signature feature and at least one type of listing text-based feature.

Abstract: Systems and methods for determining hint words that improve the accuracy of automated speech recognition (ASR) systems. Hint words are determined in the context of a user issuing voice commands in connection with a voice interface system. Terms are initially taken from most frequently occurring terms in operation of a voice interface system. For example, most frequently occurring terms that arise in electronic search queries or received commands are selected. Certain of these terms are selected as hint words, and the selected hint words are then transmitted to an ASR system to assist in translation of speech to text.

Abstract: 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.

Abstract: A method and apparatus for correcting a wrongly-translated word in a device employing speech recognition is provided herein. During operation, a device will use a second language to correct a wrongly-translated word that was wrongly translated using a first language. More particularly, after speech recognition is performed using the first language, when a user selects text to be corrected, the user will utter the speech again using the second language that differs from the first language. Both the first and the second language can be used by the device to determine a best translation of the speech.

Abstract: An information processing apparatus displays a setting screen for performing setting for a plurality of setting items for generating a product. The information processing apparatus determines whether a setting item of the plurality of setting items is selected in the setting screen. In a case where the setting item is determined to be selected in the setting screen, an image related to the selected setting item is displayed. In a case where the setting item is determined to be not selected in the setting screen, an image related to the setting item is not displayed.