Abstract: Systems and methods for product retrieval are described. One or more aspects of the systems and methods include receiving a query that includes a text description of a product associated with a brand; identifying the product based on the query by comparing the text description to a product embedding of the product, wherein the product embedding is based on a brand embedding of the brand; and displaying product information for the product in response to the query, wherein the product information includes the brand.

Abstract: The technology described herein is directed to an adaptive sparse attention pattern that is learned during fine-tuning and deployed in a machine-learning model. In aspects, a row or a column in an attention matrix with an importance score for a task that is above a threshold importance score is identified. The important row or the column is included in an adaptive attention pattern used with a machine-learning model having a self-attention operation. In response to an input, a task-specific inference is generated for the input using the machine-learning model with the adaptive attention pattern.

Abstract: Certain embodiments involve a method for generating a search result. The method includes processing devices performing operations including receiving a query having a text input by a joint embedding model trained to generate an image result. Training the joint embedding model includes accessing a set of images and textual information. Training further includes encoding the images into image feature vectors based on spatial features. Further, training includes encoding the textual information into textual feature vectors based on semantic information. Training further includes generating a set of image-text pairs based on matches between image feature vectors and textual feature vectors. Further, training includes generating a visual grounding dataset based on spatial information. Training further includes generating a set of visual-semantic joint embeddings by grounding the image-text pairs with the visual grounding dataset.

Abstract: In some embodiments, a data visualization system detects insights from a dataset and computes insight scores for respective insights. The data visualization system further computes insight type scores, from the insight scores, for insight types in the detected insights. The data visualization system determines a selected insight type for the dataset having a higher insight type score than unselected insight types and determines, for the selected insight type, a set of selected insights that have higher insight scores than unselected insights. The data visualization system determines insight visualizations for the set of selected insights and generates, for inclusion in a user interface of the data visualization system, selectable interface elements configured for invoking an editing tool for updating the determined insight visualizations from the dataset. The selectable interface elements are arranged in the user interface according to the insight scores of the set of selected insights.

Abstract: A multimodal recommendation identification system analyzes data describing a sequence of past content item interactions to generate a recommendation for a content item for a user. An indication of the recommended content item is provided to a website hosting system or recommendation system so that the recommended content item is displayed or otherwise presented to the user. The multimodal recommendation identification system identifies a content item to recommend to the user by generating an encoding that encodes identifiers of the sequence of content items the user has interacted with and generating encodings that encode multimodal information for content items in the sequence of content items the user has interacted with. An aggregated information encoding for a user based on these encodings and a system analyzes the content item sequence encoding and interaction between the content item sequence encoding and the multiple modality encodings to generate the aggregated information encoding.

Abstract: The present disclosure relates to a retrieval method including: generating a graph representing a set of users, items, and queries; generating clusters from the media items; generating embeddings for each cluster from embeddings of the items within the corresponding cluster; generating augmented query embeddings for each cluster from the embedding of the corresponding cluster and query embeddings of the queries; inputting the cluster embeddings and the augmented query embeddings to a layer of a graph convolutional network (GCN) to determine user embeddings of the users; inputting the embedding of the given user and a query embedding of the given query to a layer of the GCN to determine a user-specific query embedding; generating a score for each of the items based on the item embeddings and the user-specific query embedding; and presenting the items having the score exceeding a threshold.

Abstract: The technology described includes methods for pretraining a document encoder model based on multimodal self cross-attention. One method includes receiving image data that encodes a set of pretraining documents. A set of sentences is extracted from the image data. A bounding box for each sentence is generated. For each sentence, a set of predicted features is generated by using an encoder machine-learning model. The encoder model performs cross-attention between a set of masked-textual features for the sentence and a set of masked-visual features for the sentence. The set of masked-textual features is based on a masking function and the sentence. The set of masked-visual features is based on the masking function and the corresponding bounding box. A document-encoder model is pretrained based on the set of predicted features for each sentence and pretraining tasks. The pretraining tasks includes masked sentence modeling, visual contrastive learning, or visual-language alignment.

Abstract: Embodiments of the technology described herein describe a machine classifier capable of continually learning new classes through a continual few-shot learning approach. A natural language processing (NLP) machine classifier may initially be trained to identify a plurality of other classes through a conventional training process. In order to learn a new class, natural-language training data for a new class is generated. The training data for the new class may be few-shot training data. The training also uses synthetic training data that represents each of the plurality of other classes. The synthetic training data may be generated through a model inversion of the original classifier. The synthetic training data and the natural-language training data are used to retrain the NLP classifier to identify text in the plurality of other classes and the new class using.

Abstract: Systems, methods, and non-transitory computer-readable media are disclosed for generating generate explanatory paths for column annotations determined using a knowledge graph and a deep representation learning model. For instance, the disclosed systems can utilize a knowledge graph to generate an explanatory path for a column label determination from a deep representation learning model. For example, the disclosed systems can identify a column and determine a label for the column using a knowledge graph (e.g., a representation of a knowledge graph) that includes encodings of columns, column features, relational edges, and candidate labels. Then, the disclosed systems can determine a set of candidate paths between the column and the determined label for the column within the knowledge graph. Moreover, the disclosed systems can generate an explanatory path by ranking and selecting paths from the set of candidate paths using a greedy ranking and/or diversified ranking approach.

Abstract: A method of determining efficacy of a dataset includes receiving data from a data source, wherein the data comprises a plurality of fields of unknown efficacy; mapping the data based on a plurality of data quality metrics and based on attributes of the plurality of fields wherein meta-features for the data are obtained; predicting a value for each of the plurality of data quality metrics using a ML model that takes the meta-features as input, wherein the value indicates whether a corresponding data quality metric is suitable for measuring efficacy of the fields; selecting a data quality metric based on the value, wherein the data quality metric measures an efficacy of the fields; and monitoring the efficacy of the fields in the data received from the data source based on the data quality metric.

Abstract: Digital content search techniques are described that overcome the challenges found in conventional sequence-based techniques through use of a query-aware sequential search. In one example, a search query is received and sequence input data is obtained based on the search query. The sequence input data describes a sequence of digital content and respective search queries. Embedding data is generated based on the sequence input data using an embedding module of a machine-learning model. The embedding module includes a query-aware embedding layer that generates embeddings of the sequence of digital content and respective search queries. A search result is generated referencing at least one item of digital content by processing the embedding data using at least one layer of the machine-learning model.

Abstract: Digital content search techniques are described. In one example, the techniques are incorporated as part of a multi-head self-attention module of a transformer using machine learning. A localized self-attention module, for instance, is incorporated as part of the multi-head self-attention module that applies local constraints to the sequence. This is performable in a variety of ways. In a first instance, a model-based local encoder is used, examples of which include a fixed-depth recurrent neural network (RNN) and a convolutional network. In a second instance, a masking-based local encoder is used, examples of which include use of a fixed window, Gaussian initialization, and an adaptive predictor.

Abstract: The present disclosure relates to systems, methods, and non-transitory computer readable media that generate a target classifier for a target domain via domain adaptation using a source classifier learned on a source domain. For instance, in one or more embodiments, the disclosed systems utilize an embedding model, a target classifier, and a source classifier to analyze sets of target samples and generate classification probabilities for the target samples based on the analysis. In some cases, the disclosed systems utilize the classification probabilities to modify the parameters of the target classifier via adaptive adversarial inference. In some implementations, the disclosed systems further utilize the classification probabilities to modify the parameters of the embedding model via contrastive category-wise matching. Thus, in some cases, the disclosed systems utilize the target classifier with the modified parameters to generate classifications for digital data from the target domain.

Abstract: The technology described herein is directed to a reinforcement learning based framework for training a natural media agent to learn a rendering policy without human supervision or labeled datasets. The reinforcement learning based framework feeds the natural media agent a training dataset to implicitly learn the rendering policy by exploring a canvas and minimizing a loss function. Once trained, the natural media agent can be applied to any reference image to generate a series (or sequence) of continuous-valued primitive graphic actions, e.g., sequence of painting strokes, that when rendered by a synthetic rendering environment on a canvas, reproduce an identical or transformed version of the reference image subject to limitations of an action space and the learned rendering policy.

Abstract: Systems, methods, and non-transitory computer-readable media are disclosed for dynamically determining schema labels for columns regardless of information availability within the columns. For example, the disclosed systems can identify a column that contains an arbitrary amount of information (e.g., a header-only column, a cell-only column, or a whole column). Additionally, the disclosed systems can generate a vector embedding for an arbitrary input column by selectively using a header neural network and/or a cell neural network based on whether the column includes a header label and/or whether the column includes a populated column cell. Furthermore, the disclosed systems can compare the column vector embedding to schema vector embeddings of candidate schema labels in a d-dimensional space to determine a schema label for the column.

Abstract: A domain alignment technique for cross-domain object detection tasks is introduced. During a preliminary pretraining phase, an object detection model is pretrained to detect objects in images associated with a source domain using a source dataset of images associated with the source domain. After completing the pretraining phase, a domain adaptation phase is performed using the source dataset and a target dataset to adapt the pretrained object detection model to detect objects in images associated with the target domain. The domain adaptation phase may involve the use of various domain alignment modules that, for example, perform multi-scale pixel/path alignment based on input feature maps or perform instance-level alignment based on input region proposals.

Abstract: The present disclosure relates to systems, non-transitory computer-readable media, and methods for generating automatic suggestions to effectively modify the organization of an ingested data collection without destruction of the underlying raw data. In particular, in one or more embodiments, the disclosed systems utilize multiple machine learning models in sequence to determine likelihoods that the organizational structure of an ingested data collection should be modified in various ways. In response to generating these likelihoods, the disclosed systems generate corresponding automatic suggestions to modify the organization of the ingested data collection. In response to a detected selection of one or more of the automatic suggestions, the disclosed systems read data out of the ingested data collection in accordance with the selected automatic suggestions to effectively modify the organization of the ingested data collection.

Abstract: Adapting a machine learning model to process data that differs from training data used to configure the model for a specified objective is described. A domain adaptation system trains the model to process new domain data that differs from a training data domain by using the model to generate a feature representation for the new domain data, which describes different content types included in the new domain data. The domain adaptation system then generates a probability distribution for each discrete region of the new domain data, which describes a likelihood of the region including different content described by the feature representation. The probability distribution is compared to ground truth information for the new domain data to determine a loss function, which is used to refine model parameters. After determining that model outputs achieve a threshold similarity to the ground truth information, the model is output as a domain-agnostic model.

Abstract: One example method involves operations for a processing device that include receiving, by a machine learning model trained to generate a search result, a search query for a text input. The machine learning model is trained by receiving pre-training data that includes multiple documents. Pre-training the machine learning model by generating, using an encoder, feature embeddings for each of the documents included in the pre-training data. The feature embeddings are generated by applying a masking function to visual and textual features in the documents. Training the machine learning model also includes generating, using the feature embeddings, output features for the documents by concatenating the feature embeddings and applying a non-linear mapping to the feature embeddings. Training the machine learning model further includes applying a linear classifier to the output features. Additionally, operations include generating, for display, a search result using the machine learning model based on the input.

Abstract: Techniques are disclosed for the generation of adversarial training data through sequence perturbation, for a deep learning network to perform event sequence analysis. A methodology implementing the techniques according to an embodiment includes applying a long short-term memory attention model to an input data sequence to generate discriminative sequence periods and attention weights associated with the discriminative sequence periods. The attention weights are generated to indicate the relative importance of data in those discriminative sequence periods. The method further includes generating perturbed data sequences based on the discriminative sequence periods and the attention weights. The generation of the perturbed data sequences employs selective filtering or conservative adversarial training, to preserve perceptual similarity between the input data sequence and the perturbed data sequences.