Automated Video Switching via Audio-Visual Subject Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing live streaming technologies require manual intervention for video switching, lacking the capability for unmanned switching, which is labor-intensive and inefficient.
Innovation Solution
A signal processing device and system that utilize image and audio processing to identify and coordinate subject images and sound sources, enabling automated selection and switching of video feeds based on coincidence between subjects and sound sources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If manual switching of video streams is used, then labor requirements are high, but switching accuracy and viewer suitability are maintained through human judgment
Solution Approach 1:
The system enables self-service automation by having the video switching system automatically detect sound sources, identify corresponding subjects in video feeds, and switch displays without human intervention. The controller autonomously processes audio signals, determines sound source positions, matches them with subject images, and controls video output based on predetermined rules, making the system serve itself rather than requiring manual operation.
Solution Approach 2:
The patent replaces the mechanical/manual switching system with an automated electronic control system. Instead of manual selection of video streams, the system uses audio processing to detect sound sources, electronically identifies corresponding video feeds through subject recognition, and automatically switches displays using a controller, substituting human mechanical operation with electronic automation.
2Productivity
If automated switching without subject-sound source coordination is used, then labor is reduced, but switching accuracy and viewer suitability deteriorate
Solution Approach 1:
The system implements feedback by continuously monitoring audio signals, detecting sound source positions, comparing them with subject locations in video feeds, and adjusting video switching decisions based on this information. The controller uses the detected sound source data to feedback control which video stream should be displayed, ensuring accurate matching between sound sources and corresponding subjects while maintaining high switching efficiency.
Solution Approach 2:
The system achieves multi-functionality by integrating audio processing, subject recognition, position determination, and video switching control into a single automated system. The controller performs multiple functions simultaneously: processing audio signals to identify sound sources, determining their positions, matching them with subject images in video feeds, and controlling video output selection, thereby maintaining high accuracy across multiple operational tasks.
3Loss of time
If multiple video feeds are processed manually, then video selection quality is maintained, but time consumption and operational complexity increase
Solution Approach 1:
The system applies preliminary action by pre-establishing the correspondence relationships between sound sources and subject images in video feeds. The controller is pre-configured with the understanding of which subjects appear in which video streams, allowing it to quickly determine the appropriate video feed to display when a sound source is detected, thereby reducing switching time without requiring complex real-time analysis of multiple feeds.
Data Source
AI summary
A signal processing device includes an input interface, an image processor, an audio processor, and a controller. The input interface receives signals of a video and an audio acquired concurrently in a space where subjects exist. The image processor recognizes subject images in the video, to determine a first type of area where each subject exists. The audio processor recognizes sound sources in the audio, to determine a second type of area where each sound source exists in the space, independently of the first type of area. The controller uses the first and second types of areas to judge coincidence or non-coincidence between a position of the each subject and a position of the each sound source, to determine a combination of a subject and a sound source whose positions coincide with each other. The controller selectively determines the subject image to be output that corresponds to the combination.