Systems and methods for classifying images (e.g., ads) are described. An image is accessed. Optical character recognition is performed on at least a first portion of the image. Image recognition is performed via a convolutional neural network on at least a second portion of the image. At least one class for the image is automatically identified, via a fully connected neural network, based on one or more predictions, each of the one or more predictions being based on both the optical character recognition and the image recognition. Finally, the at least one class identified for the image is output.

Described herein are technologies that provide an element of security related to file system operations. Individual nodes in a file system, such as a directory or a file, can be associated with information that describes how to handle letter case when a file name included in a file system operation request is used to locate a file in the file system. For example, a case sensitive designation associated with a directory can require a case sensitive match between a file name included in a request and a file name included in the directory, in order to perform the requested file system operation. In another example, a case preferring designation associated with a directory first checks for a case sensitive match between file names. If a case sensitive match does not exist, then a case insensitive match between the file names can be used to perform the requested file system operation.

Described herein are technologies that provide an element of security related to file system operations. Individual nodes in a file system, such as a directory or a file, can be associated with information that describes how to handle letter case when a file name included in a file system operation request is used to locate a file in the file system. For example, a case sensitive designation associated with a directory can require a case sensitive match between a file name included in a request and a file name included in the directory, in order to perform the requested file system operation. In another example, a case preferring designation associated with a directory first checks for a case sensitive match between file names. If a case sensitive match does not exist, then a case insensitive match between the file names can be used to perform the requested file system operation.

Systems and methods are provided to enable a participant who is unable to attend all or a portion of an electronic conference to receive a machine-generated summary of the portion of the electronic conference that were missed, such as due to an intermittent network issue, and/or portions of interest when a participant is unable to attend the electronic conference while it is occurring. For example, an artificial intelligence (AI) agent, such as a neural network, may determine when a subscribed subject is being discussed and generate a summary for presentation to an absent participant. Similarly, an intermittent participant may be provided with a text and/or audio summary of the content missed during a connection drop and allow the participant to quickly catch up on the conference content missed and rejoin the electronic conference.

Systems and methods are provided to enable a participant who is unable to attend all or a portion of an electronic conference to receive a machine-generated summary of the portion of the electronic conference that were missed, such as due to an intermittent network issue, and/or portions of interest when a participant is unable to attend the electronic conference while it is occurring. For example, an artificial intelligence (AI) agent, such as a neural network, may determine when a subscribed subject is being discussed and generate a summary for presentation to an absent participant. Similarly, an intermittent participant may be provided with a text and/or audio summary of the content missed during a connection drop and allow the participant to quickly catch up on the conference content missed and rejoin the electronic conference.

A set of partial words is received. At least one partial word in the set of partial words is completed. The set of partial words with the at least one completed partial word is run through a trained deep neural network, the trained deep neural network inferring a word embedding associated with an unfinished word in the set of partial words. An inferred word is determined based on the inferred word embedding associated with the unfinished word. A sentence may be output, which includes at least the completed partial word and the inferred word.

A set of partial words is received. At least one partial word in the set of partial words is completed. The set of partial words with the at least one completed partial word is run through a trained deep neural network, the trained deep neural network inferring a word embedding associated with an unfinished word in the set of partial words. An inferred word is determined based on the inferred word embedding associated with the unfinished word. A sentence may be output, which includes at least the completed partial word and the inferred word.

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.

Method for displaying a message summary by analyzing the message to identify a sending institution and a message type. A message cluster is determined from the sending institution and repository of messages for multiple users. Extracted items can be identified in the message using the message type. Then the message summary can be generated using the extracted items, the message cluster, and the message structure. These summaries can be used to efficiently summarize a large volume of messages compactly.

Method for displaying a message summary by analyzing the message to identify a sending institution and a message type. A message cluster is determined from the sending institution and repository of messages for multiple users. Extracted items can be identified in the message using the message type. Then the message summary can be generated using the extracted items, the message cluster, and the message structure. These summaries can be used to efficiently summarize a large volume of messages compactly.