Abstract: Techniques for improving image processing related to item deliveries are described. In an example, a computer system receives an image showing a drop-off of an item, the item associated with a delivery to a delivery location. The computer system inputs the image to a first artificial intelligence (AI) model. The computer system receives first data comprising an indication of whether the drop-off is correct from the first AI model. The computer system causes a presentation of the indication at a device associated with the delivery of the item to the delivery location.

Abstract: Systems and methods are provided for detecting discrepancies in information that is displayed to a user across different webpages of an online shopping system. User interaction with the online shopping system is simulated in one example to collect data relating to webpages of the online shopping system. The data is analyzed to identify relevant semantic blocks of data and compare the data in those blocks across webpages. In an additional example, source code for the webpages is analyzed using machine-learning systems, which are targeted based on the results of the user interaction simulation to identify problematic source code and correct associated issues.

Abstract: Techniques for improving image processing related to item deliveries are described. In an example, a computer system receives image data showing a portion of a delivery location. The computer system determines an artificial intelligence (AI) model associated with the delivery location. The computer system inputs the image data to the AI model. The computer system receives an indication of whether the portion corresponds to a correct drop-off location and causes a presentation about the indication to be provided at a device.

Abstract: Aspects of connected phrase search queries and titles are described. In one embodiment, a phrase in a search string may be identified with reference to a compilation data store. The compilation data store may include various collaborative, taxonomy, and brand name data, for example. After a phrase is identified in the search string, a search query may be generated based on the phrase. For example, a catalog of items may be queried to generate a query result, and the query result may be parsed for display to a user. In various embodiments, a plurality of phrases may be identified in a search string, and a plurality of search queries may be generated based on combinations of the phrases. In this manner, more search results may be obtained, especially for constrained search strings. The identification of phrases may also be applied to generate variations on titles, in certain embodiments.

Abstract: Techniques for determining and providing a preferred layout of a network page may be provided. For example, the system may receive a query with a keyword, access data regarding previous users' interactions associated with the results of the keyword, determine a category associated with the interactions, and provide a display (e.g., on the original network page) for that particular category. In some examples, the system may provide generic network pages and item- or category-specific network pages associated with a particular layout or presentation characteristics. The system can identify the presentation characteristics of the category-specific network page and incorporate them with the generic network page, without redirecting the user to the category-specific network page.

Abstract: A classifier training system trains unified classifiers for categorizing videos representing different categories of a category graph. The unified classifiers unify the outputs of a number of separate initial classifiers trained from disparate subsets of a training set of media items. The training process takes into account the relationships that exist between the various categories of the category graph by relating scores associated with related categories, thus enhancing the accuracy of the unified classifiers.

Abstract: A method for determining a video summary from a video sequence including a time sequence of video frames, comprising: defining a global feature vector representing the entire video sequence; selecting a plurality of subsets of the video frames; extracting a frame feature vector for each video frame in the selected subsets of video frames; defining a set of basis functions, wherein each basis function is associated with the frame feature vectors for the video frames in a particular subset of video frames; using a data processor to automatically determine a sparse combination of the basis functions representing the global feature vector; determining a summary set of video frames responsive to the sparse combination of the basis functions; and forming the video summary responsive to the summary set of video frames.

Abstract: A method for identifying a set of key frames from a video sequence including a time sequence of video frames, the method executed at least in part by a data processor, comprising: selecting a set of video frames from the video sequence; identifying a plurality of visually homogeneous regions from each of the selected video frames; defining a set of basis functions, wherein each basis function is associated with a different visually homogeneous region; determining a feature vector for each of the selected video frames; representing each of the determined feature vectors as a sparse combination of the basis functions; for each of the determined feature vectors, determining a sparse set of video frames that contain the visually homogeneous regions corresponding to the basis functions included in the corresponding sparse combination of the basis functions; and analyzing the sparse sets of video frames to identify the set of key frames.

Abstract: A method for identifying a set of key frames from a video sequence including a time sequence of video frames, the method executed at least in part by a data processor, comprising: selecting a set of video frames from the video sequence; identifying a plurality of visually homogeneous regions from each of the selected video frames; defining a set of basis functions, wherein each basis function is associated with a different visually homogeneous region; determining a feature vector for each of the selected video frames; representing each of the determined feature vectors as a sparse combination of the basis functions; for each of the determined feature vectors, determining a sparse set of video frames that contain the visually homogeneous regions corresponding to the basis functions included in the corresponding sparse combination of the basis functions; and analyzing the sparse sets of video frames to identify the set of key frames.

Abstract: A method for determining a video summary from a video sequence including a time sequence of video frames, comprising: defining a global feature vector representing the entire video sequence; selecting a plurality of subsets of the video frames; extracting a frame feature vector for each video frame in the selected subsets of video frames; defining a set of basis functions, wherein each basis function is associated with the frame feature vectors for the video frames in a particular subset of video frames; using a data processor to automatically determine a sparse combination of the basis functions representing the global feature vector; determining a summary set of video frames responsive to the sparse combination of the basis functions; and forming the video summary responsive to the summary set of video frames.