Abstract: In methods for training a natural language generation (NLG) model using a processor a document-level machine translation (MT) model is provided by training an MT model to receive as input, token sequences in a first language, and to generate as output, token sequences in a second language. An augmented document-level MT model is provided by training the document-level MT model to receive as input, paired language-independent structured data and token sequences in the first language, and to generate as output, token sequences in the second language. The augmented document-level MT model is trained to receive as input, language-independent structured data, and to generate as output, token sequences in the second language.

Abstract: Am image retrieval system includes: a neural network (NN) module configured to generate local features based on an input image; an iterative attention module configured to, via T iterations, generate an ordered set of super features in the input image based on the local features, where T is an integer greater than 1; and a selection module configured to select a second image from a plurality of images in an image database based on the second image having a second ordered set of super features that most closely match the ordered set of super features in the input image, where the super features in the set of super features do not include redundant local features of the input image.

Abstract: A method is disclosed for estimating a trajectory of an object on a map given a sequence of traces for the moving object. Each trace of the object including information defining a position measured at a given time for the object, as well as information as to an area of accuracy around the measured position. The method processes pairs of successive traces, corresponding to two positions successive in time in the sequence of measured positions for the moving object. For each trace of a pair of successive traces, the method defines road segments on the map within the area of accuracy of the trace. For each road segment within the area of accuracy of a first trace of a pair of traces and each road segment within the area of accuracy of the second trace of the pair, the method determines at least one candidate path between the two road segments. A neural network and a neural graph model are used to compute the most probable sequence of candidate paths to estimate the trajectory of the object on the map.

Abstract: An interaction module includes: a first text-image interaction module configured to generate a vector representation of a first text-image pair based on an encoded representation of a reference image and an encoded representation of a text modifier, the reference image and the text modifier received from a computing device. A second text-image interaction module is configured to generate a vector representation of a second text-image pair based on the encoded representation of the text modifier and an encoded representation of a candidate target image. A compatibility module is configured to compute, based on the vector representation of the first text-image pair and the vector representation of the second text-image pair, a compatibility score for a triplet including the reference image, the text modifier, and the candidate target image. A ranking module is configured to rank a set of candidate target images including the candidate target image by compatibility scores.

Abstract: A pose estimation training system includes: a first model configured to generate a first 6 degrees of freedom (DoF) pose of a first camera that captured a first image from a first domain; a second model configured to generate a second 6 DoF pose of a second camera that captured a second image from a second domain, where the second domain is different than the first domain; a discriminator module configured to, based on first and second outputs from the first and second encoder modules, generate a discriminator output indicative of whether the first and second images are from the same domain; and a training control module configured to, based on the discriminator output, selectively adjust at least one weight value shared by the first model and the second model.

Abstract: A method of semi-supervised learning includes inputting an image; generating a weak augmentation version and a strong augmentation version of the inputted image; predicting a class of the weak augmentation version of the inputted image; determining if the predicted class of the weak augmentation version of the inputted image is confident; using a pseudo-label to train a model using the strong augmentation version of the inputted image when the predicted class of the weak augmentation version of the selected image is confident; and using a self-supervised loss based on deep clustering to train a model using the strong augmentation version of the selected image when the predicted class of the weak augmentation version of the selected image is not confident.

Abstract: A speed estimation system includes: a detection module configured to determine bounding boxes of an object moving on a surface in images, respectively, captured using a camera; a solver module configured to, based on the bounding boxes, determine a homography of the surface by solving an optimization problem, where the solver module is not trained; and a speed module configured to, using the homography, determine a speed that the object is moving on the surface.

Abstract: The present invention discloses a method of detecting an abusing act in an online service, the method including: a use act information collecting operation in which an apparatus for detecting an abusing act collects use act information including the kinds and an order of one or more use acts performed by a user; an image configuring operation in which the kinds of the one or more use acts performed by the user are arranged in a time order and an image is configured by setting the kinds of predetermined use acts to a first axis and setting the order of the use acts to a second axis; and an abusing act detecting operation in which whether the use act of the user is an abusing act is detected by using the image.

Abstract: A system for generating whole body poses includes: a body regression module configured to generate a first pose of a body of an animal in an input image by regressing from a stored body anchor pose; a face regression module configured to generate a second pose of a face of the animal in the input image by regressing from a stored face anchor pose; an extremity regression module configured to generate a third pose of an extremity of the animal in the input image by regressing from a stored extremity anchor pose; and a pose module configured to generate a whole body pose of the animal in the input image based on the first pose, the second pose, and the third pose.

Abstract: A speed estimation system includes: a detection module configured to: detect an object on a surface in an image captured using a camera; and generate a bounding box around the object; a Jacobian module configured to generate a Jacobian for the object based on the bounding box; and a speed module configured to determine a speed that the object is traveling on the surface based on the Jacobian.

Abstract: Disclosed are an electronic apparatus for dynamic note matching (DNM) and an operating method thereof, the method including acquiring a first section sequence by reducing a first sequence extracted from an input signal based on at least one first section in which the respective values are successively arranged; acquiring a second section sequence reduced from a pre-stored second sequence based on at least one second section in which the respective values are successively arranged; and calculating a similarity between the first section sequence and the second section sequence.

Abstract: Disclosed is a method and system for detecting a duplicate document using vector quantization. A duplicate document detection method may include acquiring, by processing circuitry, a respective vector expression for each of a plurality of documents using a similarity model, the similarity model being trained to output similar vector expressions for semantically similar documents, generating a key by performing a vector quantization on the respective vector expression, the key including a binary character string, and detecting a duplicate document from among the plurality of documents using the key.

Abstract: There is disclosed a method for using a computer to enable correction of misclassified labels in a database. The computer initially applies a dataset (including labels pointing respectively to categories) to a first classifier, which includes a first loss function. Pursuant to the initial application, the computer determines that one or more labels have been misclassified. Responsive to such determination, the computer changes the first loss function to a second loss function to form a second classifier including the second loss function. The computer then applies the dataset to the second classifier for enabling correction of the one or more misclassified labels.

Abstract: A robot system includes a selection module configured to select a stored demonstration for a robot from a database of stored demonstrations for different tasks of the robot; an encoder module of an attention model, the encoder module configured to determine a similarity value reflecting a similarity between: a user input demonstration for the robot; and the stored demonstration for the robot; and an indicator module configured to indicate whether the stored demonstration is the same as the user input demonstration and belongs to the same task based on the similarity value.

Abstract: An automatic scroll control method for synchronizing positions of an original text and a translation includes displaying, by a processor of a terminal device, the original text on a first display and displaying the translation corresponding to the original text on a second display; and when the original text or the translation is scrolled by a drag operation of a user, synchronizing the display positions of the original text and the translation with each other.

Abstract: A computer-implemented method for performing few-shot imitation is disclosed. The method comprises obtaining at least one set of training data, wherein each set of training data is associated with a task and comprises (i) one of samples of rewards and a reward function, (ii) one of samples of state transitions and a transition distribution, and (iii) a set of first demonstrations, training a policy network embodied in an agent using reinforcement learning by inputting at least one set of first demonstrations of the at least one set of training data into the policy network, and by maximizing a risk measure or an average return over the at least one set of first demonstrations of the at least one set of training data based on respective one or more reward functions or respective samples of rewards, obtaining a set of second demonstrations associated with a new task, and inputting the set of second demonstrations and an observation of a state into the trained policy network for performing the new task.

Abstract: A method includes: initializing a list of token embeddings, each of the token embeddings corresponding to a tokenized word from text in a corpus of text; generating a graph for a group of consecutive words s from the text, said graph including binary relations between pairs of tokenized words of the group of consecutive words; selecting the token embeddings representing the words of the group of consecutive words from the list of token embeddings; computing a tensor of binary relations as the product between a matrix of the selected token embeddings and a tensor representing discourse relations, the computed tensor representing the binary relations between the pairs of tokenized words; computing a loss using the computed tensor; optimizing the list of token embeddings using the computed loss. The above may be repeated until the computed loss is within a predetermined range.

Abstract: A training system includes: a first training dataset including first entries, wherein each of the first entries includes: a first sentence; a second sentence; and an indicator of a relationship between the first and second sentences; a training module configured to: generate a second dataset including second entries based on the first entries, respectively, wherein each of the second entries includes: the first sentence of one of the first entries; the second sentence of the one of the first entries; a first surface realization corresponding to first facts regarding the first sentence; the indicator of the one of the first entries; and train a model using the second dataset and store the model in memory.

Abstract: A distributed transcoding method includes splitting image content into segments of a fixed interval; allocating the split segments to a plurality of workers and performing parallel transcoding; and concatenating the transcoded segments and merging same into one encoded file.

Abstract: Provided are a computer system for transmitting audio content to realize a user-customized being-there and a method thereof. The computer system may be configured to detect audio files that are generated for a plurality of objects at a venue, respectively, and metadata including spatial features that are set for the objects at the venue, respectively, and to transmit the audio files and the metadata for a user. An electronic device of the user may realize a being-there at the venue by rendering the audio files based on the spatial features in the metadata. That is, the user may feel a user-customized being-there as if the user directly listens to audio signals generated from corresponding objects at a venue in which the objects are provided.