Microsoft Technology Licensing, LLC

Benjamin T Grover

Ghassan G Salloum

Oscar M P Koenders

Elliott O Downing

Deepak Kumar

Gaurav Bhandari

Brian Anger

Ashima Jain

A digital image is automatically cropped to fit within a desired frame. The cropping is based on one or more of two identified portions of the image. One of the portions is an all-subjects portion that includes all the identified subjects of a particular type in the image. The other portion is an attention portion that identifies an intended focus of the image. An attempt to crop the image to include both of these portions is made, and if unsuccessful then an attempt to crop the image to include at least the all-subjects portion is made. If neither of these attempts is successful, then the image is cropped to include one or more, but less than all, of the identified subjects of the particular type in the image.

Automated image cropping to include particular subjects

Techniques for application quality testing time predictions are described that provide customized processing time predictions for application testing. Testing times for various stages of testing are collected during application testing. The collected testing data may be stored in association with various attribute values ascertained for the applications. The collected timing data is used to predict processing times for applications that match defined profiles. The profiles may be defined to include selected attributes and values used to identify applications having corresponding attributes and values. The collected timing data may be processed to derive predictions for processing times on a profile-by-profile basis. When a new application is submitted, the application is matched to a particular profile based on attribute values possessed by the application. Then, processing time predictions associated with the matching profile may be obtained, assigned to the application, and output for presentation to a developer of the application.

Application quality testing time predictions

An application for which approval is requested is identified and multiple automated tests are applied to the application in groups of automated tests. Each of the groups of automated tests includes multiple ones of the multiple automated tests. If one or more automated tests in a group of automated tests returns an inconclusive result, then a manual check is initiated for the application based on the one or more automated tests that returned the inconclusive result. If one or more automated tests in a group, or a manual test applied in the manual check, returns a fail result then an indication that the application is rejected is returned, the indication that the application is rejected including an identification of why the application is rejected. If none of the multiple automated tests returns a fail result, then a manual testing phase is initiated.

Managing automated and manual application testing

Growing community of inventors

Issaquah, WA, United States of America

Benjamin T Grover