Abstract: Systems and methods are disclosed to more efficiently and effectively search for videos. A video sharing system may receive a video from a user, may extract features from the received video and may store the extract features for the video. Moreover, based on a prespecified re-encoding scheme, the video sharing system re-encodes the received video. For example, the video re-encoding may be performed by generating a set of video segments from video data of the received video such that each video segment is independently playable by a media player. The video sharing platform then stores the re-encoded video including information for each video segment of the set of video segments generated during the re-encoding process of the received video. The re-encoded videos can then be used, in real-time, to dynamically generate search result videos that includes snipes from multiple videos that match a given search query.

Abstract: A head mounted display (HMD) identifies viewing trends for a video based on multiple viewings of the video by users having different user characteristics. The analysis identifies viewing trends based on user characteristics, e.g., the age and gender of the user. When a subsequent user is viewing the video, a production & post production module receives the user's profile information and identifies the predicted viewing location of the user at particular times during the video using the identified trends based on where previous users/viewers, who have a profile or characteristics similar to the user, were looking at the particular times during the video.

Abstract: A production system determines which areas or portions of a video file (e.g., for a scene) are static and which areas contain motion. Instead of steaming redundant image data for the static areas of the video, the production only sends image data for or updates the areas of each frame that contain motion to minimize an amount of data being streamed to a head mounted display (HMD) without compromising image quality.

Abstract: Variable image data reduction is applied to at least a subset of each frame of a video file. The variable image data reduction includes reducing data by one or more techniques, such as compression, decimation, distortion, and so forth, across the frame by a different degree or amount based on a viewing location of the user. Thus, the amount of data sent to an encoder for delivery to a client device (e.g., head mounted display (HMD) or other computing device) is lowered by prioritizing image quality for the viewing location of the user while one or more data reducing techniques are applied to the remainder of the frame.

Abstract: A production system determines which areas or portions of a video file (e.g., for a scene) are static and which areas contain motion. Instead of steaming redundant image data for the static areas of the video, the production only sends image data for or updates the areas of each frame that contain motion to minimize an amount of data being streamed to a head mounted display (HMD) without compromising image quality.

Abstract: Variable image data reduction is applied to at least a subset of each frame of a video file. The variable image data reduction includes reducing data by one or more techniques, such as compression, decimation, distortion, and so forth, across the frame by a different degree or amount based on a viewing location of the user. Thus, the amount of data sent to an encoder for delivery to a client device (e.g., head mounted display (HMD) or other computing device) is lowered by prioritizing image quality for the viewing location of the user while one or more data reducing techniques are applied to the remainder of the frame.

Abstract: A head mounted display (HMD) identifies viewing trends for a video based on multiple viewings of the video by users having different user characteristics. The analysis identifies viewing trends based on user characteristics, e.g., the age and gender of the user. When a subsequent user is viewing the video, a production & post production module receives the user's profile information and identifies the predicted viewing location of the user at particular times during the video using the identified trends based on where previous users/viewers, who have a profile or characteristics similar to the user, were looking at the particular times during the video.

Abstract: One embodiment uses a computer system to define areas with moving image information of interest in a sequence of digital moving image frames. The sequence of digital moving image frames has at least three frames, a first frame, a second frame and a third frame, in time sequence. A set of points identifies the areas with moving image information of interest on the first frame. In the third frame, the areas with moving image information of interest are identified with the set of points. Each point in the set of points has an explicitly defined location in the first frame and an explicitly defined location in the third frame. The location of each point in the first frame corresponds to the same areas with moving image information of interest as each point's location in the second and third frames. For example if one of the points is located at the tip of a runner's left toe in the first frame, then that point will be located at the tip of the runner's left toe in both the second and third frames.