Bookmarking and Seeking Tool for Online Videos
Rahimi Motem, Siamak (2017)
Rahimi Motem, Siamak
2017
Information Technology
Tieto- ja sähkötekniikan tiedekunta - Faculty of Computing and Electrical Engineering
This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Hyväksymispäivämäärä
2017-06-07
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:tty-201705261550
https://urn.fi/URN:NBN:fi:tty-201705261550
Tiivistelmä
At 2014, 66% of internet traffic was related to video content [1]. This number and everyday experience shows that by improving handheld device capabilities, social networks and internet speed, the video content which has been seen and posted is taking up most internet traffic. As a result, this thesis focuses on improving the user experience with videos in two supplementary features: Bookmarking videos and enhanced seeking.
There have been many cases, such as CCTV and medical cases, where making a video summary and video synopsis does not serve the purpose and the whole video must be available. However, the user is only interested in certain moments in the video. Usually in these cases either a video summary is generated along the main video, interesting moments in the video is kept as a note, or the user finds it manually by making seeking forward and backward.
Video bookmarking, which means keeping the original video and makes a list of interesting moments in the video, so that the user can seek toward them by selecting them solves this issue. The bookmarks are standardized JSON objects in a JSON array that can be added, deleted or modified. In their simplest form, they have a relative start time, duration and a caption.
Having bookmarks available in the cases mentioned above, user behavior can be predicted. The user is highly likely to request a seek for a bookmarked moment. By caching the video content, which has the bookmarked content, the user does not need to wait for buffering to see the video playing from the seek target. Currently, the user must wait for buffering. It has a major impact in cases such as CCTV and medical cases, where different cameras have recorded a scene from different angles and a seek action must seek all the video content, at the same time.
In this thesis, an application has been developed as proof of concept which has met both requirements. It has been developed over an existing application, which is used for treatment of epilepsy by using automated seizure detection.
There have been many cases, such as CCTV and medical cases, where making a video summary and video synopsis does not serve the purpose and the whole video must be available. However, the user is only interested in certain moments in the video. Usually in these cases either a video summary is generated along the main video, interesting moments in the video is kept as a note, or the user finds it manually by making seeking forward and backward.
Video bookmarking, which means keeping the original video and makes a list of interesting moments in the video, so that the user can seek toward them by selecting them solves this issue. The bookmarks are standardized JSON objects in a JSON array that can be added, deleted or modified. In their simplest form, they have a relative start time, duration and a caption.
Having bookmarks available in the cases mentioned above, user behavior can be predicted. The user is highly likely to request a seek for a bookmarked moment. By caching the video content, which has the bookmarked content, the user does not need to wait for buffering to see the video playing from the seek target. Currently, the user must wait for buffering. It has a major impact in cases such as CCTV and medical cases, where different cameras have recorded a scene from different angles and a seek action must seek all the video content, at the same time.
In this thesis, an application has been developed as proof of concept which has met both requirements. It has been developed over an existing application, which is used for treatment of epilepsy by using automated seizure detection.