Google Inc.

Pittsburgh Pattern Recognition, Inc.

Michael Nechyba

Michael Andrew Sipe

Henry Will Schneiderman

Steven Ross

Yong Zhao

Uriel Gerardo Rodriguez

Louis D Brandy

Brian Colonna

Brian Patrick Williams

Andrew C Gallagher

Utsav Prabhu

Hang Chu

Terrence Paul McCartney, Jr

Terrance Paul McCartney, Jr

A method for aligning a translation of original caption data with an audio portion of a video is provided. The method involves identifying original caption data for the video that includes caption character strings, identifying translated language caption data for the video that includes translated character strings associated with audio portion of the video, and mapping caption sentence fragments generated from the caption character strings to corresponding translated sentence fragments generated from the translated character strings based on timing associated with the original caption data and the translated language caption data. The method further involves estimating time intervals for individual caption sentence fragments using timing information corresponding to individual caption character strings, assigning time intervals to individual translated sentence fragments based on estimated time intervals of the individual caption sentence fragments, generating a set of translated sentences using consecutive translated sentence fragments, and aligning the set of translated sentences with the audio portion of the video using assigned time intervals of individual translated sentence fragments from corresponding translated sentences.

Automated voice translation dubbing for prerecorded videos

A method for aligning a translation of original caption data with an audio portion of a video is provided. The method includes identifying, by a processing device, original caption data for a video that includes a plurality of caption character strings. The processing device identifies speech recognition data that includes a plurality of generated character strings and associated timing information for each generated character string. The processing device maps the plurality of caption character strings to the plurality of generated character strings using assigned values indicative of semantic similarities between character strings. The processing device assigns timing information to the individual caption character strings based on timing information of mapped individual generated character strings. The processing device aligns a translation of the original caption data with the audio portion of the video using assigned timing information of the individual caption character strings.

Automated voice translation dubbing for prerecorded video

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for partitioning videos. In one aspect, a method includes obtaining a partition of a video into one or more shots. Features are generated for each shot, including visual features and audio features. The generated features for each shot are provided as input to a partitioning neural network that is configured to process the generated features to generate a partitioning neural network output. The partition of the video into one or more chapters is determined based on the partitioning neural network output, where a chapter is a sequence of consecutive shots that are determined to be taken at one or more locations that are semantically related.

Partitioning videos

A transformed video and a source video may be synchronized according to implementations disclosed herein to provide tag information to the device receiving the transformed version of the video. A synchronization signal may be computed on the source video and the transformed video using a statistic such as mean pixel intensity. The synchronization signals computed for the transformed video and source video may be compared to determine a transformed video reference point location for the requested tag information. The requested tag information may be provided to the device receiving the transformed version of the video.

Synchronizing videos with frame-based metadata using video content

Facial recognition can be used to determine whether or not a user has access privileges to a device such as a smartphone. It may be possible to spoof an authenticated user's image using a picture of the user. Implementations disclosed herein utilize time of flight and/or reflectivity measurements of an authenticated user to compare to such values obtained for an image of an object. If the time of flight distance and/or reflectivity distance for the object match a measured distance, then the person (i.e., object) may be granted access to the device.

Differentiating real faces from representations

An example method includes capturing, by a camera of a computing device, an image including at least a face of a user, calculating a face template of the face of the user in the image, and analyzing the face template to determine whether the face includes at least one of a removable facial feature that decreases a level of distinctiveness between two faces and a non-removable facial feature that decreases a level of distinctiveness between two faces. When the face includes the removable facial feature, the method further includes outputting a notification for the user to remove the removable facial feature. When the face includes the non-removable facial feature, the method further includes adjusting a first similarity score threshold to a second similarity score threshold.

Facial feature detection

An example method includes capturing, by an image capture device of a computing device, an image of a face of a user. The method further includes detecting, by the computing device, whether a distance between the computing device and an object represented by at least a portion of the image is less than a threshold distance, and, when the detected distance is less than a threshold distance, denying authentication to the user with respect to accessing one or more functionalities controlled by the computing device, where the authentication is denied independent of performing facial recognition based at least in part on the captured image.

Facial recognition

An example method includes receiving a first image and a second image of a face of a user, where one or both images have been granted a match by facial recognition. The method further includes detecting a liveness gesture based on at least one of a yaw angle of the second image relative to the first image and a pitch angle of the second image relative to the first image, where the yaw angle corresponds to a transition along a horizontal axis, and where the pitch angle corresponds to a transition along a vertical axis. The method further includes generating a liveness score based on a yaw angle magnitude and/or a pitch angle magnitude, comparing the liveness score to a threshold value, and determining, based on the comparison, whether to deny authentication to the user with respect to accessing one or more functionalities controlled by the computing device.

An example method includes capturing, by a camera of a mobile computing device, an image, determining whether the image includes a representation of at least a portion of a face, and, when the image includes the representation of at least the portion of the face, analyzing characteristics of the image. The characteristics include at least one of a tonal distribution of the image that is associated with a darkness-based mapping of a plurality of pixels of the image, and a plurality of spatial frequencies of the image that are associated with a visual transition between adjacent pixels of the image. The method further includes classifying, by the mobile computing device, a quality of the image based at least in part on the analyzed characteristics of the image.

Facial image quality assessment

In general, aspects of the present disclosure are directed to techniques for adjusting the threshold similarity score to match the facial features of an authorized user in a facial recognition system. Instead of a uniform threshold similarity score, representations of authorized faces enrolled in the computing device may be assigned a custom threshold similarity score based at least in part on the distinctiveness of the facial features in the representations of the authorized faces. A computing device can determine a distinctiveness of facial features of an enrolling user, and can determine a threshold similarity score associated with the facial features of the enrolling user based at least in part on the distinctiveness of the facial features of the enrolling user.

Facial recognition similarity threshold adjustment

An example method includes receiving first and second images of a face of a user, where one or both images have been granted a match by facial recognition. The method includes identifying at least one facial landmark in the first image and at least one corresponding facial landmark in the second image, and extracting a first sub-image from the first image, where the first sub-image includes a representation of the at least one facial landmark. The method includes extracting a second sub-image from the second image, where the second sub-image includes a representation of the at least one corresponding facial landmark, detecting a facial gesture by determining whether a sufficient difference exists between the second sub-image and first sub-image to indicate the facial gesture, and determining, based on detecting the facial gesture, whether to deny authentication to the user with respect to accessing functionalities controlled by the computing device.

Growing community of inventors

Pittsburgh, PA, United States of America

Michael Nechyba