Display Apparatus Dynamic Content Recognition Period Adjustment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing display apparatuses face high resource and cost burdens due to frequent server queries for content recognition, especially when detecting changes in viewed content, leading to inefficient data traffic and processing loads.
Innovation Solution
A display apparatus that dynamically adjusts the content recognition period based on content type and viewing history, using a combination of local and server-based automatic content recognition (ACR) to minimize server load and enhance accuracy, by calculating probabilities of content change and predicting next content for optimized fingerprint reception.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If the display apparatus frequently sends queries to the server to rapidly detect content changes, then the content recognition speed is improved, but the resource consumption and cost increase significantly
Solution Approach 1:
The system performs preliminary actions by pre-processing video frames to generate fingerprints and pre-judging whether content changes occur before actually querying the server. This filtering mechanism prepares the data in advance and eliminates unnecessary server communications, reducing resource consumption while maintaining recognition speed.
Solution Approach 2:
The content recognition process is segmented into multiple stages: fingerprint generation from video frames, local pre-judging of content changes, and conditional server querying. This segmentation allows the system to handle only necessary cases at the resource-intensive server level, optimizing the balance between speed and resource usage.
2Measurement precision
If the display apparatus frequently sends queries to the server to rapidly detect content changes, then the content recognition accuracy is improved, but the data traffic and server load increase
Solution Approach 1:
The system extracts only the essential fingerprint features from video frames and transmits only when content changes are detected locally. This extraction principle removes unnecessary data transmission by sending only relevant recognition information to the server, reducing data traffic while maintaining recognition accuracy.
Solution Approach 2:
The fingerprint analysis system acts as an intermediary between local content detection and server recognition. It processes and filters content change information before server communication, serving as a mediator that reduces unnecessary data traffic while ensuring accurate content recognition through coordinated local-server processing.
3Loss of energy
If the display apparatus uses local fingerprint matching, then the server load is reduced, but the content recognition capability is limited without server support
Solution Approach 1:
The system performs preliminary fingerprint generation and local matching before server communication. This preliminary action allows the display apparatus to handle routine recognition independently, reducing server load while maintaining reliable content identification for common cases without requiring server support.
Data Source
Figure 1~2A
Figure 2B~3
Figure 4A
AI summary
A display apparatus and a method for recognizing a content thereof are provided. The display apparatus includes a display; a memory storing a fingerprint which is obtained based on a characteristic of a content, and information regarding the content corresponding to the fingerprint; a transceiver; and at least one processor configured to: obtain a fingerprint based on a characteristic of a screen of a content currently reproduced on the display, search presence/absence of a stored fingerprint matching the obtained fingerprint in the memory, identify, based on a result of the searching, whether to transmit a query comprising the obtained fingerprint to a server through the transceiver to request information on the currently reproduced content, identify a type of the currently reproduced content using a data recognition model that uses the obtained fingerprint, and change a content recognition period based on the identified type of the currently reproduced content.