Disclosed herein are methods, systems, and apparatus, including computer programs encoded on computer storage media, for electronically stamping a document. One of the methods include receiving an electronic stamping instruction, where the electronic stamping instruction comprises a to-be-stamped document and a stamping type. In response to determining that a format of the to-be-stamped document is a predetermined document format and the stamping type is a first stamping type, a first to-be-stamped area of the to-be-stamped document is determined. An electronic stamp corresponding to the to-be-stamped document is identified using an encryption algorithm interface. A first electronically stamped document is generated and include the electronic stamp in the first to-be-stamped area.

Systems and methods are provided for determining a media cost equivalent to attribute to the appearance of a sponsor's logo or other object in a media item, such as an image or video. This determination may be based on factors such as clarity, prominence, visibility, size, or placement of a sponsor logo or other object related to the sponsor within at least one media item. One or more machine learning models may be utilized to automatically identify objects within, and otherwise analyze, image data in order to determine sponsor value.

A data generation apparatus includes: an acquisition unit configured to acquire an image of an object to be inspected including a defect, and a region of the image that includes the defect; a correction unit configured to correct the region acquired by the acquisition unit by expanding an outer edge of the region so that the number of pixels included in the region is increased by a predetermined amount; and a generation unit configured to generate learning data by associating the region corrected by the correction unit with the image.

Provided are an image recognition apparatus, an image recognition method, and a program for enabling recognition of many kinds of objects with high precision. An overall recognition unit executes, for at least one given object, a process of recognizing the position of the object in an image. A partial image extraction unit extracts, from the image, a partial image which is a part of the image associated with the recognized position. A partial recognition unit executes a process of recognizing what is one or more objects represented by the partial image, the one or more objects including an object other than the given object the position of which is recognized.

Systems and methods are provided for analyzing images or video using computer vision. Data comprising real time or near real time information or historical information is retrieved that is associated with a sporting event at a physical location. A time segment is identified of a display device at the physical location for acquisition. The display device is configurable to present visual sponsorship data during the time segment for an assigned sponsor. It is determined that one or more rules are satisfied by the data. An indication is transmitted that the first rule is satisfied to a computing device of a sponsor. A bid or valuation is generated based at least on the first rule being satisfied. A request to acquire the time segment is received from the computing device of the sponsor, and the display device at the physical location is caused to present visual sponsorship data for the sponsor during the time segment.

A document processing system is configured to identify, for each accessed electronic document in a first set of multiple electronic documents, a set of identified multi-word phrases determined to be in ordered text information in the accessed electronic document, each multi-word phrase of the set of identified multi-word phrases including adjacent words in the ordered text information; and determine, for each accessed electronic document in the first set of multiple electronic documents, a selected document type from the first set of document types based at least on an analysis of the set of identified multi-word phrases with respect to multi-word-phrase characteristics identified by a first definition and associated with each document type in a first set of document types associated with a first document-set type.

Disclosed herein are techniques related to automated generation of regular expressions. In some embodiments, a regular expression generator may receive input data comprising one or more character sequences. The regular expression generator may convert character sequences into a sets of regular expression codes and/or span data structures. The regular expression generator may identify a longest common subsequence shared by the sets of regular expression codes and/or spans, and may generate a regular expression based upon the longest common subsequence.

Aspects of the present disclosure relate to managing the upload of media items to cloud storage. A model can be configured to distinguish between confidential and non-confidential media items. A media item to be uploaded to a cloud storage can be analyzed using the model to determine whether the media item is confidential. In response to determining that the media item is confidential, an upload control action can be issued on the media item.

A text correction method and apparatus can take advantage of a greatly reduced number of error-ground truth pairs to train a deep learning model. To generate these error-ground truth pairs, different characters in a ground truth word are replaced with a symbol, not appearing in any ground truth words, to generate error words which are paired with that ground truth word to provide error-ground truth word pairs. This process may be repeated for all ground truth words for which training is to be performed. In embodiments, pairs of characters in a ground truth word may be replaced with a symbol to generate the error words which are paired with that ground truth word to provide error-ground truth word pairs. Again, this process may be repeated for all ground truth words for which training is to be performed.

Disclosed herein are techniques related to automated generation of regular expressions. In some embodiments, a regular expression generator may receive input data comprising one or more character sequences. The regular expression generator may convert character sequences into a sets of regular expression codes and/or span data structures. The regular expression generator may identify a longest common subsequence shared by the sets of regular expression codes and/or spans, and may generate a regular expression based upon the longest common subsequence.