AI Content Extraction for Reliable Multimedia Data Broadcasting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing broadcast technologies face challenges in providing extensive coverage and deep indoor penetration, especially during network congestion or in areas with poor reception, leading to incomplete message delivery and accessibility issues.
Innovation Solution
A method involving AI-based content extraction algorithms to decompose multimedia data into smaller, non-audio, non-video objects, which are then broadcast with enhanced protection, allowing reception devices to reconstruct the content using local reconstruction data, even in degraded conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If broadcast technologies transmit complete multimedia data, then content quality is maintained, but data volume increases causing network congestion and reduced delivery reliability
Solution Approach 1:
The patent segments complete multimedia data into essential objects and non-essential objects. Only essential objects are transmitted through the broadcast network, while non-essential objects are obtained locally from a database. This segmentation reduces transmitted data volume while maintaining content completeness, directly resolving the contradiction between data volume and delivery reliability.
Solution Approach 2:
The patent extracts and transmits only the most critical elements (essential objects) of the multimedia content through the broadcast network. Non-critical elements are extracted from local databases at receiving devices. This extraction approach minimizes network data volume while ensuring reliable delivery of core content.
2Area of stationary object
If broadcast signals are transmitted with high power to achieve deep indoor penetration, then coverage area increases, but energy consumption increases and network congestion worsens
Solution Approach 1:
By segmenting content into essential and non-essential objects, the patent enables transmission of smaller data packets that require lower transmission power. Receiving devices obtain non-essential objects locally without requiring additional broadcast power, thus expanding effective coverage area while reducing energy consumption.
3Speed
If network processes all broadcast requests simultaneously, then message delivery speed is maximized, but network congestion increases during high traffic periods
Solution Approach 1:
The patent segments the message delivery process into two parallel paths: essential objects are delivered quickly through broadcast network, while non-essential objects are retrieved locally from databases. This segmentation enables simultaneous processing without congestion, as the bulk of data retrieval occurs locally rather than through the network.
Solution Approach 2:
Non-essential objects are pre-stored in local databases at receiving devices before broadcast events. When a broadcast occurs, devices immediately access pre-stored data locally without requiring network processing, thereby maintaining fast delivery speed while reducing network processing load during high traffic periods.
4Loss of information
If complete multimedia content is transmitted to ensure content accuracy, then information completeness is maintained, but indoor penetration capability decreases due to signal weakness
Solution Approach 1:
The patent segments content into essential objects (transmitted via broadcast) and non-essential objects (obtained locally). This segmentation enables complete content reconstruction at receiving devices even with weak indoor signals, as only critical minimal data needs to penetrate indoors while non-critical data comes from local storage.
Data Source
Figure 1
Figure 2
Figure 3a~3b
AI summary
The present invention relates to a method for broadcasting data, the method being implemented at a transmitting device and comprising: - extracting a set of objects from the data by using a content extraction method; and - broadcasting the set of objects by using a broadcasting technique; wherein the data comprises audio or video data associated with an audio or video content; wherein the set of objects incompletely represents the audio or video content; wherein the content extraction method comprises an Artificial Intelligence, Al, -based extraction algorithm.