User-Adaptive Volume Selection for Virtual Assistants
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Virtual assistants often generate audio responses that are either too quiet or too loud for the context in which they are provided, leading to indiscernible or disruptive interactions.
Innovation Solution
A volume selection model is calibrated based on user input, using machine learning features such as sound characteristics, ambient noise, and device settings to adapt the volume level of audio responses to the specific context.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If the virtual assistant uses a fixed volume level for audio responses, then the device complexity is reduced, but the adaptability to different contexts deteriorates
Solution Approach 1:
The system performs preliminary analysis of the audio input characteristics (loudness, distance indicators, background noise) before generating the audio response, and pre-determines the appropriate volume level based on this analysis. This allows the system to adapt to different contexts without requiring complex real-time adjustments during response delivery.
Solution Approach 2:
The system uses feedback from the user's audio input characteristics (such as loudness level, distance from device based on echo analysis, and background noise levels) to automatically adjust the volume of the audio response. This feedback mechanism enables contextual adaptation while keeping the overall system relatively simple.
2Reliability
If the virtual assistant uses the last-used volume level, then the ease of operation is maintained, but the reliability of appropriate volume selection deteriorates
Solution Approach 1:
The system automatically analyzes the context of the current interaction (user's voice loudness, distance from device, background noise) and self-determines the appropriate volume level without requiring the user to manually adjust settings. This maintains ease of operation while significantly improving the reliability of volume appropriateness for each specific context.
3Reliability
If the audio response volume is increased to ensure discernibility, then the sound level is improved, but the harmful disruption to the environment increases
Solution Approach 1:
The system adjusts the audio response volume locally based on the specific characteristics of each interaction context. By analyzing indicators such as user distance from the device and background noise levels, the system applies different volume levels appropriate to each local situation, ensuring discernibility when needed while minimizing disruption in quiet environments.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Techniques for user-adaptive volume selection are disclosed. A system receives audio input that includes a user request to a virtual assistant. The system determines one or more sound characteristics corresponding to the audio input, and applies the sound characteristic(s) to a volume selection model. The volume selection model selects a volume level based on the sound characteristic(s). The system plays an audio response to the user's request at the selected volume level. Subsequently, the system receives user input to modify output volume to a different volume level. Based on the user input, the system calibrates the volume selection model. When the system receives another audio input, the system analyzes the other audio input to determine one or more sound characteristics of the other audio input. The system uses the calibrated volume selection model to select a volume level for an audio response to the other audio input.