Video Labeling via Event Information Extraction and Plot Trigger Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current video labeling methods are inefficient due to high workload in tracking playback amounts, limited availability of operation data for all videos, and reliance on comment information with low reference values, resulting in low labeling efficiency.
Innovation Solution
A method where a server extracts event information from video frames, forms event information flows, and captures clips that meet plot trigger conditions to generate plot labeling tags, improving video labeling efficiency by identifying key content directly from image content.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If video labeling is performed according to playback amount, then wonderful clips can be identified, but enormous workload is required to count playback amount at each moment, resulting in low video labeling efficiency
Solution Approach 1:
The patent extracts event information from video frames to identify wonderful clips, removing the need to count playback amounts at each moment. By extracting key event information (such as plot triggers, character actions, scene changes) directly from video content, the system achieves accurate wonderful clip identification without the enormous computational workload of tracking every playback moment.
Solution Approach 2:
The patent replaces the mechanical counting process of playback amounts with an intelligent event extraction system. Instead of mechanically tracking and counting playback data point-by-point, the system uses event information extraction and analysis to automatically identify wonderful clips, significantly reducing computational complexity while maintaining or improving identification accuracy.
2Ease of operation
If video labeling is performed according to operation data, then wonderful clips can be labeled, but some videos lack corresponding operation data, resulting in low video labeling efficiency
Solution Approach 1:
The patent creates a universal video labeling system that works with all video types regardless of whether operation data is available. By using event information extraction from video frames as the core mechanism, the system can label wonderful clips in any video (game videos, movies, tutorials, etc.) without being limited by the availability of specific operation data, achieving broad applicability and consistent efficiency.
Solution Approach 2:
The patent enables videos to label their own wonderful clips through self-contained event information extraction. Instead of relying on external operation data that may not be available, the system extracts event information directly from the video content itself, allowing any video to be labeled independently without requiring additional external data sources.
3Loss of information
If video labeling is performed according to comment information, then wonderful clips can be extracted, but the wonderful clip is limited by reference value of comment information, resulting in low video labeling efficiency when reference value is low
Solution Approach 1:
The patent replaces the mechanical reliance on comment information with an intelligent event extraction system. Instead of depending on user comments as the sole source for identifying wonderful clips, the system extracts event information directly from video frames and analyzes plot structures, character actions, and scene changes to automatically identify wonderful clips, eliminating the limitation of low reference value in comment data.
Solution Approach 2:
The patent introduces event information as an intermediary between video content and wonderful clip identification. Rather than directly using comment information (which may have low reference value), the system uses extracted event information as a mediator to objectively identify wonderful clips based on actual video content, plot structures, and event sequences, providing a more reliable and efficient labeling mechanism.
4Measurement precision
If event information is extracted from video frames and event information flows are formed, then plot trigger conditions can be captured to identify wonderful clips, but the processing complexity increases
Solution Approach 1:
The patent segments the video labeling process into distinct modules: event information extraction from frames, event information flow formation, plot trigger condition matching, and wonderful clip identification. By dividing the complex processing into sequential, independent segments, the system manages complexity effectively while maintaining high identification accuracy through systematic event analysis.
Solution Approach 2:
The patent performs preliminary extraction and organization of event information from video frames before conducting plot trigger condition analysis. By pre-processing video content to extract and structure event information in advance, the system reduces the complexity of subsequent wonderful clip identification, as the foundation of event data is already prepared and organized for efficient pattern matching.
Data Source
AI summary
This application provides a video labeling method performed by a server, and the method includes: receiving a video extraction instruction transmitted by a terminal, and obtaining a to-be-extracted video according to the video extraction instruction; extracting event information from video frames of the to-be-extracted video; forming at least one event information flow by using the event information; capturing, based on the at least one event information flow, at least one first clip that meets a plot trigger condition from the to-be-extracted video and obtaining a plot labeling tag of the at least one first clip; and transmitting the at least one first clip and the corresponding plot labeling tag to the terminal, wherein the terminal displays the at least one first clip and the corresponding plot labeling tag in a preset display region of a display interface in which the to-be-extracted video is displayed.