Conference Virtual Assistant Audio Isolation for Private Queries
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
There is a need to integrate video conferencing and virtual assistants effectively, allowing participants to engage with virtual assistants without disrupting the conference or issuing confusing queries.
Innovation Solution
Systems and methods for activating a virtual assistant during a conference by identifying a command in the audio input, stopping audio transmission to other participants, and performing actions based on user queries through the virtual assistant.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If a virtual assistant is activated during a video conference, then user convenience is improved, but conference integrity is worsened due to audio disruption
Solution Approach 1:
The audio stream is segmented into conference audio and virtual assistant audio channels. The system separates the audio input stream into distinct portions: conference audio that continues uninterrupted and virtual assistant audio that is isolated when the VA is active. This segmentation allows both functions to operate simultaneously without mutual interference, resolving the contradiction between user convenience and conference integrity.
Solution Approach 2:
An audio mixing system acts as an intermediary between the video conference audio stream and the virtual assistant audio stream. The mixing system combines these audio streams in a way that allows both to coexist without disruption - the conference audio continues normally while the virtual assistant audio is isolated when needed, thus mediating between the conflicting requirements of user convenience and conference integrity.
2Reliability
If audio transmission is stopped when virtual assistant is activated, then virtual assistant performance is improved, but user convenience is worsened due to audio disruption
Solution Approach 1:
The audio transmission is segmented into separate channels for conference audio and virtual assistant audio. Instead of stopping all audio transmission, the system segments the audio stream and continues transmitting conference audio while isolating the virtual assistant audio portion when the VA is active. This maintains user convenience while ensuring virtual assistant performance.
Solution Approach 2:
The audio transmission system dynamically adjusts its behavior based on the operational state. When the virtual assistant is inactive, full audio transmission occurs for conference communication. When the virtual assistant is activated, the system dynamically switches to a segmented transmission mode where conference audio continues and virtual assistant audio is isolated. This dynamic adaptation resolves the contradiction between virtual assistant performance and user convenience.
3Productivity
If wake word detection is performed continuously, then virtual assistant responsiveness is improved, but energy consumption is worsened
Solution Approach 1:
Instead of continuous wake word detection, the system employs periodic detection with adjustable intervals. The wake word detection operates at different frequencies based on system state - more frequent when needed and less frequent during stable states. This periodic action maintains virtual assistant responsiveness while significantly reducing energy consumption compared to continuous monitoring.
Solution Approach 2:
The system changes the detection parameter (sampling frequency) dynamically based on operational requirements. When the virtual assistant needs to be highly responsive, the wake word detection operates at higher frequency. When responsiveness requirements are lower, the detection frequency is reduced. This parameter change approach resolves the contradiction between productivity and energy consumption.
Data Source
AI summary
Systems and methods are provided for enabling virtual assistant interaction during a conference. A conference is initiated between a first computing device and at least a second computing device, and audio input is received at an audio input device, wherein the audio input is received during the conference and the audio input device is in communication with the first computing device. The audio input is transmitted to the second computing device, and a command for activating a virtual assistant is identified in the audio input. In response to identifying a command, the virtual assistant is activated, and the transmission of the audio input to at least the second computing device is automatically stopped. A query is received at the audio input device, and an action, based on the query, is performed via the virtual assistant.