Subtitle Rendering Based on Reading Pace
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current closed captioning and subtitling methods fail to ensure that captioned text can be read within the time frame of the displayed scene, often requiring users to rewind and replay content multiple times, and do not adequately account for user language proficiency or provide suitable translations.
Innovation Solution
A system that automatically summarizes captioned text based on user language proficiency and reading pace, adjusts playback speed, and rewrites captions to ensure readability within the scene's timeframe, using machine learning and AI to personalize the text and provide options for user approval or rejection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If closed captioning provides word-for-word speech transcript synchronized frame-by-frame, then the user can read along with the media asset, but the amount of captioned text requires far greater time and cannot be read while the associated scene is displayed
Solution Approach 1:
The patent segments the captioned text into multiple versions with different levels of detail (e.g., full transcript vs. summarized version). The system allows users to select the appropriate segmentation level based on their reading pace and the scene duration, enabling them to read captions completely within the time frame without requiring the full word-for-word transcript.
Solution Approach 2:
The patent applies partial action by providing users with the option to read only the essential information from captions rather than requiring complete word-for-word transcription. The system can selectively display summarized captions that contain the key information needed to understand the scene, reducing the time required while maintaining comprehension.
2Ease of operation
If captions are displayed at normal speed, then the user can read them, but users must rewind and replay multiple times to read the full captions before the scene changes
Solution Approach 1:
The patent implements dynamic adjustment of caption display parameters based on real-time user behavior detection. When the system detects that a user is rewinding or taking time to read captions, it dynamically modifies the caption delivery speed or provides summarized versions that can be read more quickly, eliminating the need for repeated rewinding and replaying.
Solution Approach 2:
The system incorporates feedback mechanisms that monitor user interaction with captions and adjust the caption display accordingly. If users spend excessive time on a particular scene or repeatedly rewind, the system receives feedback and adapts by providing summarized captions or adjusting the reading pace to match the user's actual reading speed, thereby reducing the need for multiple viewings.
3Adaptability or versatility
If subtitles translate foreign language dialog, then the media asset can be watched by viewers who do not understand the language, but the translation may not be suitable for users with different language proficiency levels
Solution Approach 1:
The patent applies local quality by customizing the translation and caption content to match the specific language proficiency level of each user. Instead of providing a single translation version, the system offers multiple versions (e.g., literal translation, simplified translation, or even original audio with transcripts) and selects the appropriate one based on the user's detected language proficiency, ensuring the translation is both accurate and accessible.
Solution Approach 2:
The system changes parameters such as translation complexity, vocabulary level, and sentence structure based on the user's language proficiency profile. By adjusting these parameters dynamically, the system can adapt the same foreign language dialogue into different caption versions that suit users ranging from beginners to advanced language learners, maintaining both accuracy and accessibility.
Data Source
AI summary
Systems and methods for summarizing captions, configuring playback speed, and rewriting the caption file for a media asset are disclosed. The system determines whether to display the original captions or a summarized version of the captions, which are based on user's language proficiency level, reading pace, and historical data, and can be generated either on-demand or automatically when rewinds and pauses are detected. The caption file which includes the original captions can be rewritten. The system determines whether to stream a caption or a rewritten file to a media device based on user or system selections. In the absence of a caption file, or when the caption file cannot be summarized, the playback speed of the media asset is slowed down to provide additional reading time to the user.