Abstract: Provided are systems and methods for automatic video processing that employ machine learning models to process input video and understand user video content in a semantic and cultural context. This recognition enables the processing system to recognize interesting temporal events, and build narrative video sequences automatically, for example, by linking or interleaving temporal events or other content with film-based categorizations. In further embodiments, the implementation of the processing system is adapted to mobile computing platforms which can be distributed as an “app” within various app stores. In various example, the mobile apps turn everyday users into professional videographers. In further embodiments, music selection and dialog based editing can likewise be automated via machine learning models to create dynamic and professional quality video segments.

Abstract: Provided are systems and methods for automatic video processing that employ machine learning models to process input video and understand user video content in a semantic and cultural context. This recognition enables the processing system to recognize interesting temporal events, and build narrative video sequences automatically, for example, by linking or interleaving temporal events or other content with film-based categorizations. In further embodiments, the implementation of the processing system is adapted to mobile computing platforms which can be distributed as an “app” within various app stores. In various example, the mobile apps turn everyday users into professional videographers. In further embodiments, music selection and dialog based editing can likewise be automated via machine learning models to create dynamic and interest professional quality video segments.

Abstract: In some implementations, a 360-degree camera includes two wide-angle lenses that provide a spherical view of a scene. The 360-degree camera is configured to be connected to a computing device (e.g., a smart phone) for rendering the captured images. The user interface provides options to set camera orientations during playback in order to present a selected orientation in the view while the video is displayed. Additionally, specific orientations may be set in some frames and the video processor provides for smooth transitions from the orientation in one frame to the orientation in the next specified frame. This way, the user may follow the action on a particular sequence during playback. The resulting video playback may be saved as a movie. The user interface also provides the option to set user-defined landmarks and follow these landmarks, for a user-configured amount of time, during playback.

Abstract: In some implementations, a 360-degree camera includes two wide-angle lenses that provide a spherical view of a scene. The 360-degree camera is configured to be connected to a computing device (e.g., a smart phone) for rendering the captured images. In some implementations, the camera lenses in the 360-degree camera are wide-angle lenses (e.g., 208°) and corrections are performed to the projection of the 3D video data into a planar view to adjust the fisheye effect. A fisheye mapping function is selected that improves the distribution of sampling rays to present a natural view without blurred areas.

Abstract: Systems and methods for editing an image file include a server and at least one client device of the server including a display. An imaging module accesses from the server an image file including image content and a header, wherein the header provides information regarding the image file. An editing module receives user edits to the image content and insert information regarding the user edits into the header. The imaging module applies the user edits in an order that is determined based on a weight assigned to each user edit. The imaging module may also access, from the server, an image file including image content and a header thereof, wherein the header provides information regarding the image file and an edit decision list reflecting historical user edits to the image content. The editing module then identifies the edit decision list in the header for application by the imaging module.

Abstract: Methods, systems, and apparatus for adjusting a perspective of an image are described. Pixels from each of a plurality of rows of the image are removed, where the count of pixels removed is based on a location of the corresponding row in the image. In one example embodiment, the image is rotated prior to and after removal of the pixels based on the type of perspective alteration.

Abstract: Systems and methods for omnidirectional application navigation are provided. In example embodiments, a navigation icon associated with a first user interface context is caused to be displayed on a device. A change is sensed, by a device, of a change from the first user interface context to a second user interface context. A graphics database is accessed. From the graphics database, a graphic which is associated with the second user interface context is identified. Responsive to the change from the first user interface context to the second user interface context, the navigation icon is changed using the identified graphic from the graphics database.

Abstract: Systems and methods for menuless application navigation are provided. In example embodiments, a first user input received via an interface of the device is detected. It is determined that the first user input is within an input region on the interface of the device. In response to determining that the first user input is within the input region, the first user input is translated into a navigation action. A second user input received via the interface of the device is detected, and it is determined that the second user input is outside the input region on the interface of the device. In response to determining that the second user input is outside the input region, the user input is translated into a second action, which is different from the navigation action.

Abstract: An image capture and processing system accesses an image including image data. The image processing system receives a selection of a preset configuration. The preset configuration identifies first and second image effects, a first intensity value corresponding to the first image effect, and a second intensity value corresponding to the second image effect. The image processing system accesses a set of stored lookup tables (LUTs). Each accessed LUT in the set of stored LUTs is associated with the first and second image effects, and is associated with intensity values within a threshold proximity of the first and second intensity values. The image processing system generates a combined LUT by interpolating the accessed set of stored LUTs. The image processing system processes the image based on the combined LUT.

Abstract: An image capture and processing system accesses an image including image data. The image processing system receives a selection of a preset configuration. The preset configuration identifies first and second image effects, a first intensity value corresponding to the first image effect, and a second intensity value corresponding to the second image effect. The image processing system accesses a set of stored lookup tables (LUTs). Each accessed LUT in the set of stored LUTs is associated with the first and second image effects, and is associated with intensity values within a threshold proximity of the first and second intensity values. The image processing system generates a combined LUT by interpolating the accessed set of stored LUTs. The image processing system processes the image based on the combined LUT.

Abstract: A system and method for correcting the perspective of visual data in real-time is disclosed. A electronic device accesses live visual data through a camera associated with an electronic device. The electronic device displays the live visual data on a display associated with the electronic device. The electronic device detects tilt data from a sensor associated with the electronic device, wherein tilt data indicates that the electronic device has tilted from an original angle to a current angle. In response to detecting tilt data from the sensor associated with the electronic device, the electronic devices alters the displayed live visual data to correct a live perspective of a live video feed.

Abstract: An image capture and processing system accesses an image including image data. The image processing system receives a selection of a preset configuration. The preset configuration identifies first and second image effects, a first intensity value corresponding to the first image effect, and a second intensity value corresponding to the second image effect. The image processing system accesses a set of stored lookup tables (LUTs). Each accessed LUT in the set of stored LUTs is associated with the first and second image effects, and is associated with intensity values within a threshold proximity of the first and second intensity values. The image processing system generates a combined LUT by interpolating the accessed set of stored LUTs. The image processing system processes the image based on the combined LUT.

Abstract: Systems and methods for editing an image file include a server and at least one client device of the server including a display. An imaging module accesses from the server an image file including image content and a header, wherein the header provides information regarding the image file. An editing module receives user edits to the image content and insert information regarding the user edits into the header. The imaging module applies the user edits in an order that is determined based on a weight assigned to each user edit. The imaging module may also access, from the server, an image file including image content and a header thereof, wherein the header provides information regarding the image file and an edit decision list reflecting historical user edits to the image content. The editing module then identifies the edit decision list in the header for application by the imaging module.

Abstract: A system and method for correcting the perspective of visual data in real-time is disclosed. A electronic device accesses live visual data through a camera associated with an electronic device. The electronic device displays the live visual data on a display associated with the electronic device. The electronic device detects tilt data from a sensor associated with the electronic device, wherein tilt data indicates that the electronic device has tilted from an original angle to a current angle. In response to detecting tilt data from the sensor associated with the electronic device, the electronic devices alters the displayed live visual data to correct a live perspective of a live video feed.

Abstract: Systems and methods for editing an image file include a server and at least one client device of the server including a display. An imaging module accesses from the server an image file including image content and a header, wherein the header provides information regarding the image file. An editing module receives user edits to the image content and insert information regarding the user edits into the header. The imaging module applies the user edits in an order that is determined based on a weight assigned to each user edit. The imaging module may also access, from the server, an image file including image content and a header thereof, wherein the header provides information regarding the image file and an edit decision list reflecting historical user edits to the image content. The editing module then identifies the edit decision list in the header for application by the imaging module.