Machine learned video template usage
Techniques for video generation based on machine learned video template usage are disclosed. A plurality of videos is obtained, and video scene analysis on each video is performed. Video cuts for each video are detected, and objects within each video are identified. The identifying includes detecting a person, face, building, or vehicle. Metadata is categorized for each of the videos based on the scene analysis, the video cuts, and the objects within the videos. Template information is stored, including the categorized metadata, on each of the videos. Each video is stored as a template video along with the template information. The template information on a subset of videos is ranked. A basis video is selected based on the template information. A further video is generated based on the basis video. The further video is stored as a further template video along with the template videos.
1. A computer-implemented method for video generation comprising:
obtaining a plurality of videos, wherein video scene analysis is performed on each of the plurality of videos;
detecting video cuts for each of the plurality of videos;
identifying objects within each of the plurality of videos;
categorizing metadata for each of the plurality of videos based on the scene analysis, the video cuts, and the objects within the plurality of videos;
storing template information, including the metadata which was categorized, on each of the plurality of videos;
selecting a basis video from the plurality of videos based on the template information;
obtaining user-added text;
generating a further video based on the basis video and a second basis video, wherein the further video includes the user-added text added to one or more scenes of the further video; and
recommending one or more additional videos for inclusion in the further video, based on the obtained user-added text, wherein the further video comprises a two-layer template video, wherein the two-layer template video comprises the second basis video superimposed on the basis video.
2. The method of claim 1 wherein each video of the plurality of videos is stored as a template video along with the template information.
3. The method of claim 2 wherein one of the template videos is used as the basis video.
4. The method of claim 2 further comprising storing the further video as a further template video along with the template videos.
5. The method of claim 4 wherein the further template video includes further template video information on scene analysis, video cuts, and objects within the further video.
6. The method of claim 2 wherein the template videos are obtained through crowdsourcing.
7. The method of claim 1 wherein a portion of each video of the plurality of videos is stored as a template video along with the template information.
8. The method of claim 7 wherein the portion comprises a template video module.
9. The method of claim 1 further comprising ranking the template information on a subset of videos from the plurality of videos.
10. The method of claim 9 further comprising selecting a video template based on the template information that was ranked.
11. The method of claim 9 further comprising recommending a plurality of video templates based on the ranking of the template information.
12. The method of claim 11 wherein the basis video is selected from the plurality of video templates that were recommended.
13. The method of claim 9 wherein the ranking is based on a view count, an engagement score, a segment duration, a comparison of template information with user text input, a subject for the further video, or a classification of the subset of videos for each video from the subset of videos.
14. The method of claim 9 wherein the ranking is provided in a non-deterministic manner.
15. The method of claim 9 wherein the ranking is based on analysis of user provided video for video content and metadata associated with the user provided video.
16. The method of claim 1 wherein the generating includes augmenting the basis video with personalized video content.
17. The method of claim 1 wherein the identifying objects includes a confidence level of an object that is identified.
18. The method of claim 1 wherein the video scene analysis includes analysis of content of a scene in a video from the plurality of videos.
19. The method of claim 18 wherein the video cuts define boundaries of a scene in the video.
20. The method of claim 1 wherein the selecting the basis video is accomplished via automatically curating a subset of the plurality of videos.
21. The method of claim 1 further comprising enabling video production.
22. The method of claim 1 wherein the basis video includes a viewing screen, and wherein the second basis video is superimposed on the viewing screen from the basis video.
23. A computer program product embodied in a non-transitory computer readable medium for video generation, the computer program product comprising code which causes one or more processors to perform operations of:
obtaining a plurality of videos wherein video scene analysis is performed on each of the plurality of videos;
detecting video cuts for each of the plurality of videos;
identifying objects within each of the plurality of videos;
categorizing metadata for each of the plurality of videos based on the scene analysis, the video cuts, and the objects within the plurality of videos;
storing template information, including the metadata which was categorized, on each of the plurality of videos;
selecting a basis video from the plurality of videos based on the template information;
obtaining user-added text;
generating a further video based on the basis video and a second basis video, wherein the further video includes the user-added text added to one or more scenes of the further video; and
recommending one or more additional videos for inclusion in the further video, based on the obtained user-added text, wherein the further video comprises a two-layer template video, wherein the two-layer template video comprises the second basis video superimposed on the basis video.
24. A computer system for video generation comprising:
a memory which stores instructions;
one or more processors coupled to the memory wherein the one or more processors, when executing the instructions which are stored, are configured to:
obtain a plurality of videos wherein video scene analysis is performed on each of the plurality of videos;
detect video cuts for each of the plurality of videos;
identify objects within each of the plurality of videos;
categorize metadata for each of the plurality of videos based on the scene analysis, the video cuts, and the objects within the plurality of videos;
store template information, including the metadata which was categorized, on each of the plurality of videos;
select a basis video from the plurality of videos based on the template information;
obtain user-added text;
generate a further video based on the basis video and a second basis video, wherein the further video includes the user-added text added to one or more scenes of the further video; and
recommend one or more additional videos for inclusion in the further video, based on the obtained user-added text, wherein the further video comprises a two-layer template video, wherein the two-layer template video comprises the second basis video superimposed on the basis video.