Voice Response Device Selection Using Distance and DNN
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional IoT devices face challenges in accurately determining the intended device to respond to a wake-up voice, often waking up unintended devices due to similar distances from the user, leading to the need for re-uttering the wake-up voice to access the specific device.
Innovation Solution
A method and apparatus that determine a response device by receiving audio signals from multiple IoT devices, extracting distance information, and using a deep neural network (DNN) model to select the appropriate device based on predetermined conditions, such as max voting schemes and probability thresholds, to ensure the intended device is activated.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional distance-based wake-up determination is used, then device simplicity is maintained, but device selection accuracy deteriorates when multiple devices are at similar distances
Solution Approach 1:
The determination process is segmented into two distinct steps: first comparing distance information using max voting scheme, and if distances are similar, then applying DNN model for more sophisticated analysis. This segmentation allows the system to handle simple cases efficiently while providing accurate resolution for ambiguous cases.
Solution Approach 2:
A max voting scheme is introduced as an intermediary mechanism between simple distance comparison and complex DNN analysis. The max voting scheme processes distance information from multiple devices and provides a preliminary determination, only invoking the more complex DNN model when needed for tie-breaking similar distance cases.
2Reliability
If simple distance comparison is used, then processing speed is maintained, but user intention recognition accuracy deteriorates in critical situations
Solution Approach 1:
The determination system dynamically adjusts its processing approach based on the situation. When distance differences are clear, the system uses fast comparison. When distances are similar (critical situations), the system dynamically switches to the more accurate but slower DNN model, optimizing the trade-off between speed and accuracy for each specific case.
Solution Approach 2:
The system changes the determination parameter based on the situation: using simple distance values for clear cases and switching to DNN-processed features for critical cases where distance alone is insufficient. This parameter adaptation allows the system to maintain high reliability while minimizing processing time.
3Measurement precision
If DNN model is always applied, then device selection accuracy is improved, but processing complexity and time consumption increase
Solution Approach 1:
Instead of always applying the full DNN model, the system applies it partially - only when necessary for critical situations with similar distances. For most cases with clear distance differences, simpler comparison methods suffice, providing partial action that maintains accuracy where needed while preserving efficiency where possible.
Data Source
AI summary
Disclosed are a response device determination method and a response device determination apparatus. The method includes receiving audio signals from a plurality of devices respectively; extracting a plurality of distance information indicative of distances between the user and the plurality of devices from the audio signals respectively; and determining a response device to respond to the wake-up voice using the extracted plurality of distance information, wherein the response device is determined based on at least one of first and second steps according to a predetermined condition, wherein the first step includes comparing the extracted plurality of distance information with each other and determining the response device based on the comparison result, wherein the second step includes applying the extracted plurality of distance information to a deep neural network (DNN) model to obtain an application result and determining the response device based on the application result. Thus, in an environment where a plurality of devices recognize the same wake-up voice, the method may select a device to respond to a wake-up voice in response to a single utterance from the user so as to reflect intention of the user.