A9.com, Inc.

Amazon Technologies, Inc.

Other

Imdb.com, Inc.

Douglas Ryan Gray

Colin Jon Taylor

Xiaofan Lin

Arnab Sanat Kumar Dhua

Yu Lou

Adam Carlson

Ashutosh Vishwas Kulkarni

Sunil Ramesh

Adam Wiggen Kraft

Ismet Zeki Yalniz

Edward Hsiao

Atul Kumar

Arnab S Dhua

Zhijiang Mark Lu

Jeffrey P Bezos

Charles Benjamin Franklin Waggoner

Jeromey Russell Goetz

David Creighton Mott

Alexander L Honda

Ming Du

Danny Ryan Stephens

Anna Makris

Charles Benjamin Waggoner

Gautam Bhargava

Aishwarya Natesh

Mark Jay Nitzberg

Nityananda Jayadevaprakash

Son Dinh Tran

Shruti Sheorey

Wei-Hong Chuang

Maya Kabkab

Arnab Sunat Kumar Dhua

Colin Jay Taylor

Xinliang Zhu

Combined feature vectors may be generated to map features of two or more search queries to a common embedding space. A user may provide an initial input query and then provide a refinement query. Independent feature vectors may be generated for each of the initial input query and the refinement query, may be weighted, and then may be combined to form a combined feature vector. The combined feature vector aligns different search modalities within the common embedding space that may be executed against an index.

Multimodal search and refinement

A computing device can obtain data describing at least one document, the at least one document referencing at least one media object, wherein a portion of the at least one media object includes one or more characters. The computing device can obtain data describing the one or more characters in the at least one media object in the at least one document. The computing device can generate an updated copy of the at least one document that includes the data describing the one or more characters in the at least one media object. The computing device can present, on a display screen of the computing device and through an interface, the updated copy of the at least one document, wherein the one or more characters in the at least one media object are able to be selected or searched.

Media processing techniques for enhancing content

Various embodiments enable a computing device to perform tasks such as processing an image to recognize text or an object in an image to identify a particular product or related products associated with the text or object. In response to recognizing the text or the object as being associated with a product available for purchase from an electronic marketplace, one or more advertisements or product listings associated with the product can be displayed to the user. Accordingly, additional information for the associated product can be displayed, enabling the user to learn more about and purchase the product from the electronic marketplace through the portable computing device.

Augmented reality recommendations

An image creation and editing tool can use the data produced from training a neural network to add stylized representations of an object to an image. An object classification will correspond to an object representation, and pixel values for the object representation can be added to, or blended with, the pixel values of an image in order to add a visualization of a type of object to the image. Such an approach can be used to add stylized representations of objects to existing images or create new images based on those representations. The visualizations can be used to create patterns and textures as well, as may be used to paint or fill various regions of an image. Such patterns can enable regions to be filled where image data has been deleted, such as to remove an undesired object, in a way that appears natural for the contents of the image.

Neural network-based image manipulation

A user can select an object represented in video content in order to set a magnification level with respect to that object. A portion of the video frames containing a representation of the object is selected to maintain a presentation size of the representation corresponding to the magnification level. The selection provides for a 'smart zoom' feature enabling an object of interest, such as a face of an actor, to be used in selecting an appropriate portion of each frame to magnify, such that the magnification results in a portion of the frame being selected that includes the one or more objects of interest to the user. Pre-generated tracking data can be provided for some objects, which can enable a user to select an object and then have predetermined portion selections and magnifications applied that can provide for a smoother user experience than for dynamically-determined data.

Object tracking in zoomed video

Approaches provide for analyzing image data to determine and/or recognize text in the image data. The recognized text can be used to generate a search query that can be automatically submitted to a search engine without having to type the search query to identify a product (or related products) associated with the image. For example, a camera of a computing device can be used to capture a live camera view (or single images) an item. An application executing on the computing device (or at least in communication with the computing device) can analyze the image data of the live camera view to determine a set of keywords (e.g., identified text) based on visual features extracted from the image data. The keywords can be used to query an index of product titles, common search queries, among other indexed data to return a ranked list of search suggestions based on a relevance function. The relevance function can consider the ordering of the keywords to rank search suggestions more highly that contain the keywords having the same word order. Further, the relevance function can consider the confidence of the visual recognition of each keyword, the confidence of each search suggestion, customer impact, as well as other factors to determine the ranking of the search suggestions. The search suggestions can be further refined to ensure search results that the user will be more likely to view and/or purchase.

Visual search suggestions

A video segmentation system can be utilized to automate segmentation of digital video content. Features corresponding to visual, audio, and/or textual content of the video can be extracted from frames of the video. The extracted features of adjacent frames are compared according to a similarity measure to determine boundaries of a first set of shots or video segments distinguished by abrupt transitions. The first set of shots is analyzed according to certain heuristics to recognize a second set of shots distinguished by gradual transitions. Key frames can be extracted from the first and second set of shots, and the key frames can be used by the video segmentation system to group the first and second set of shots by scene. Additional processing can be performed to associate metadata, such as names of actors or titles of songs, with the detected scenes.

Video segmentation techniques

Various embodiments provide methods and systems for providing a recommended volume level in presentation of media content. In some embodiments, volume adjustment events made by a user and/or similar users while watching media content can be detected and automatically recorded. The media content may include a plurality of segments. A normalized volume level for at least one segment of the media content can be determined by aggregating the recorded volume adjustment events corresponding to the at least one segment of the media content. When the media content is played back on a user device, at least some embodiments cause the at least one segment of the media content to be played back at a recommended volume level determined based at least in part upon one of the normalized audio level of the corresponding segment, the audio system of the user device, or historical data and personal profile of the user.

Crowdsourced audio normalization for presenting media content

Techniques are described for providing functionality to allow a viewer of a television show to watch a 'previously on' segment of an episode of the television show and be able to watch the scenes from prior episodes referenced in the 'previously on' segment.

Expanded previously on segments

Systems and approaches are provided for optimizing self-portraiture. The background of the self-portrait can be enhanced by image registration or stitching techniques of images captured using one or more conventional cameras. Multiple standard resolution images can be stitched together to generate a panoramic or a composite image of a higher resolution. Foreground elements, such as one or more representations of users, can also be enhanced in various ways. The representations of the users can be composited to exclude undesirable elements, such as image data of one of the users extending her arm to capture the self-portrait. An ideal pose of the users can automatically be selected and other image enhancements, such as histogram optimization, brightness and contrast optimization, color-cast correction, or reduction or removal of noise, can automatically be performed to minimize user effort in capturing self-portraits.

Self-portrait enhancement techniques

A set of training images is obtained by analyzing text associated with various images to identify images likely demonstrating a visual attribute. Localization can be used to extract patches corresponding to these attributes, which can then have features or feature vectors determined to train, for example, a convolutional neural network. A query image can be received and analyzed using the trained network to determine a set of items whose images demonstrate visual similarity to the query image at least with respect to the attribute of interest. The similarity can be output from the network or determined using distances in attribute space. Content for at least a determined number of highest ranked, or most similar, items can then be provided in response to the query image.

Attribute similarity-based search

Growing community of inventors

Mountain View, CA, United States of America

Douglas Ryan Gray