Abstract: Techniques are described for playing back a portion of an image frame of media content. The playback can be based on characteristics of the viewer device and active region data indicating important areas of the image frame.

Abstract: Techniques are described for low-pass filtering image data of virtual reality content before compression based on an understanding of the relationship of the image data to a cylindrical or polygonal projection space and the resulting compression of that image data when mapped to a playback projection space.

Abstract: Disclosed are various embodiments for selecting fragments of a media item. An estimated bandwidth for a client is calculated. A confidence score for the estimated bandwidth is determined. When the confidence score falls below a threshold, the estimated bandwidth is modified. A fragment for the media item is selected using the modified estimated bandwidth.

Abstract: A technology is described for providing plotline information associated with media content. An example method may include receiving a request from a customer via an interface for plotline information stored on a data store. The plotline information may be included in media content where the plotline information may include a plurality of plotlines featured in the narrative of the media content. The plotline information may then be identified in the data store, and a graphical visualization of the plotline information included in the media content may be generated, where the graphical visualization may display the plotline information as corresponding to the narrative of the media content.

Abstract: Techniques are described for optimizing video decoder operations.

Abstract: Techniques and systems are provided for providing linear entertainment content for perspective-shiftable displays. Some of the provided techniques and systems allow for pre-existing three-dimensional assets, such as models, scene definitions, virtual lights and cameras, motion paths, etc., that were used to create an original animated piece of linear entertainment content may be re-used and mapped, if necessary, into a format that is compatible with a perspective-shiftable display in order to re-create that same entertainment content in a perspective-shiftable format. In some instances where such pre-existing content is used, the technique may also include inserting additional content in locations that are only visible to the viewer when viewed from a perspective other than the perspectives used in creating the original animated piece of linear entertainment content.

Abstract: Techniques are described for dynamically restructuring an encoded video stream to support a wide range of high-speed playback with accompanying visual feedback of the underlying video content. For a particular range of playback speeds, frames in each fragment are retained or dropped based on the encoded dependencies among the frames in the fragment and the selected playback speed. The set of frames in each fragment is analyzed to determine the optimal pattern of frames to retain or drop for a given playback speed. The time stamps of the retained frames are adjusted to achieve the proper spacing for the decoder.

Abstract: Techniques are described for optimizing memory used by a video decoder. A residual coefficient matrix including non-zero value residual coefficients of a larger parent matrix with both non-zero and zero value residual coefficients can be provided to the decoder. Residual coefficient matrix metadata can also be provided so that a modified and reduced inverse transform matrix can be selected and applied to the residual coefficient matrix.

Abstract: Disclosed are various embodiments for requesting fragments of a media item. A latency to a media distribution service and bandwidth for a client are estimated. A time to request a subsequent fragment from the media item is determined. Sources for the fragment are scored and one of the sources is selected. The fragment is requested from the selected source.

Abstract: A technology for content delivery is provided. In one example, performance of a caching network, performance of a delivery network, and customer demand are modeled. Instructions are provided for a client device on how to request content based on the modeled performance of the caching network, the modeled performance of the delivery network or the modeled customer demand.

Abstract: Disclosed are various embodiments for requesting fragments of a media item. A latency to a media distribution service and bandwidth for a client are estimated. A time to request a subsequent fragment from the media item is determined. Sources for the fragment are scored and one of the sources is selected. The fragment is requested from the selected source.

Abstract: Techniques are described for optimizing memory used by a video decoder. A residual coefficient matrix including non-zero value residual coefficients of a larger parent matrix with both non-zero and zero value residual coefficients can be provided to the decoder. Residual coefficient matrix metadata can also be provided so that a modified and reduced inverse transform matrix can be selected and applied to the residual coefficient matrix.

Abstract: Techniques for delivering content are described that vary the bit rate with which the content is delivered to achieve a consistent level of quality from the user's perspective. This is achieved through the use of quality metrics associated with content fragments that guide decision making in selecting from among the different size fragments that are available for a given segment of the content. Fragment selection attempts to optimize quality within one or more constraints.

Abstract: 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.

Abstract: Disclosed are various embodiments for a manifest generation service. A manifest generation service determines a bit rate for an on demand content stream to a client. A manifest is generated indicating a source for content at the determined bit rate. The bit rate is determined and manifests are generated repeatedly to adapt to changes in network conditions during the stream.

Abstract: 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.

Abstract: A technology for aspect ratio validation is provided. An object may be detected in a visual media presentation. The detected object may be compared to a first candidate aspect ratio object of a same class of objects as the object. The detected object may also be compared to a second candidate aspect ratio object, which may also be of a same class of objects as the detected object. A determination may be made as to which of the first and second candidate aspect ratio objects the object corresponds.

Abstract: Techniques and systems are provided for providing linear entertainment content for perspective-shiftable displays. Some of the provided techniques and systems allow for pre-existing three-dimensional assets, such as models, scene definitions, virtual lights and cameras, motion paths, etc., that were used to create an original animated piece of linear entertainment content may be re-used and mapped, if necessary, into a format that is compatible with a perspective-shiftable display in order to re-create that same entertainment content in a perspective-shiftable format. In some instances where such pre-existing content is used, the technique may also include inserting additional content in locations that are only visible to the viewer when viewed from a perspective other than the perspectives used in creating the original animated piece of linear entertainment content.

Abstract: A technology for content delivery is provided. In one example, performance of a caching network, performance of a delivery network, and customer demand are modeled. Instructions are provided for a client device on how to request content based on the modeled performance of the caching network, the modeled performance of the delivery network and the modeled customer demand.

Abstract: Technology is described for identifying a scene in video content. Video content to be played on a device may be provided. A segment of the video content that includes an occurrence of an interest event may be identified, the segment being identified based on instructions received while the video content is being provided to the device. A scene may be created with the segment based on a starting time code of the interest event. The scene with the segment may be provided for playback.