Personalized Trickplay Slider Preview Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current media players face challenges in allowing users to quickly navigate media assets during trick-play operations due to non-personalized and densely spaced previews, making it difficult for users to find desired portions of media assets.
Innovation Solution
A media player application generates personalized, high-resolution, low-density previews based on user profiles and the importance of frames to the media asset's plot, selecting one frame per scene for display during scrubbing operations, allowing users to quickly navigate to desired time points.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If dense I-frame renditions (one frame/second) are used to provide more previews during trick-play operations, then the number of previews increases, but the user interface becomes more cluttered and harder to navigate
Solution Approach 1:
The patent extracts only the most important frames from each scene to display as previews, rather than showing all frames. This selective extraction process identifies key moments in each scene and displays only those, reducing the number of previews from potentially hundreds per second to a manageable few per scene, thereby maintaining navigability while preserving essential content information
Solution Approach 2:
The patent applies different display densities to different temporal regions of the media asset. Important scenes receive focused preview display with higher quality and prominence, while less important scenes receive reduced preview density. This local differentiation allows users to quickly identify and navigate to significant portions of the content without being overwhelmed by uniform preview distribution throughout the entire asset
2Ease of manufacture
If predetermined interval previews (one preview/second, one preview/three seconds, one preview/minute) are used, then the preview spacing is uniform and simple to implement, but the previews are not personalized for a particular user or based on frame importance
Solution Approach 1:
The system performs preliminary analysis of the media asset content before playback to identify important scenes and frames. During trick-play operations, the system then displays previews selectively based on this pre-computed importance data, showing only frames from significant moments rather than uniform intervals. This preliminary preparation enables personalized, context-aware preview display that adapts to content structure and user interests
Solution Approach 2:
The preview display system dynamically adjusts which frames are shown based on multiple factors including user profile data, scene importance metrics, and current playback context. Rather than using fixed predetermined intervals, the system dynamically selects frames to display, changing the preview density and content based on what is most relevant to the user at each moment, thereby achieving both personalization and adaptability
3Reliability
If multiple I-frame playlists with different bitrate variants are implemented, then the trick-play performance improves, but the device complexity and processing requirements increase
Solution Approach 1:
The patent changes the selection parameter from bitrate-based I-frame selection to importance-based frame selection. Instead of maintaining multiple playlists with different bitrates and selecting based on technical performance metrics, the system uses a single playlist structure but selects frames for display based on their semantic importance to the content. This parameter change maintains trick-play reliability while significantly reducing system complexity by eliminating the need for multiple bitrate variant playlists
Data Source
AI summary
Systems and methods are provided for displaying personalized preview images during trick-play operations of a media asset based on user profile data and an importance of frames of the media asset. A request from a user to move a slider of a progress bar associated with a media asset is received while the media asset is being generated for display. A scene from the media asset that contains a play position indicated by the slider is identified and a frame that best matches user profile data of the user and importance data for the scene is identified. The identified frame is generated for display as an overlay over the displayed media asset.