Air conditioning controller
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional air conditioners face misrecognition issues when multiple users have similar voiceprints, leading to inaccurate air-conditioning control based on user preferences.
Innovation Solution
An air-conditioner controller that stores voiceprint and personal data for each user, using a control unit to compare detected voice data with stored voiceprints and present questions based on personal data to accurately identify users through their responses, thereby improving recognition accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If voiceprint comparison is used to identify users, then user recognition speed is improved, but recognition accuracy deteriorates when multiple users have similar voiceprints
Solution Approach 1:
The user identification process is segmented into two distinct stages: first, voiceprint-based preliminary identification to quickly narrow down candidate users; second, question-based verification to accurately confirm the user's identity. This segmentation allows the system to balance speed and accuracy by handling different aspects of identification separately.
Solution Approach 2:
Questions based on personal data serve as an intermediary verification mechanism between voiceprint matching and final user identification. When voiceprint similarity threshold is met, the system introduces question-based verification as an intermediate step to resolve ambiguities in cases of similar voiceprints, thereby improving accuracy without significantly impacting overall recognition speed.
2Device complexity
If only voiceprint data is stored for each user, then storage simplicity is improved, but user identification reliability deteriorates due to similar voiceprints
Solution Approach 1:
The system performs preliminary action by collecting and storing personal data (such as name, age, gender, occupation) during the user registration phase, before actual user identification is needed. This preliminary data collection enables the system to conduct question-based verification during identification, improving reliability without adding complexity to the storage structure.
Solution Approach 2:
The identification system uses a composite approach by combining two types of data: voiceprint data (biometric information) and personal data (demographic information). This composite data structure leverages the strengths of both data types - voiceprint for quick matching and personal data for verification - thereby improving identification reliability while maintaining storage simplicity.
3Ease of operation
If voiceprint matching threshold is lowered to accept more users, then ease of operation is improved, but recognition accuracy deteriorates due to false positives
Solution Approach 1:
The system applies partial verification action by using question-based verification only when necessary - specifically, when voiceprint similarity meets the threshold but user identity needs confirmation. This partial action approach maintains ease of operation for clear matches while improving accuracy for ambiguous cases, avoiding the need to lower or raise the threshold globally.
Data Source
Figure 1~2
Figure 3
Figure 4
AI summary
An air-conditioner controller (3) that controls an operation of an air conditioner includes: a storage unit (15) that stores, for each user, voiceprint data that is data indicating a voiceprint of a user and personal data that is data indicating a feature of the user; and a control unit (14) that compares voice data that is data of a user's voice detected in an air-conditioning area that is an area subject to air-conditioning control of the air conditioner with the voiceprint data, and, when there is voiceprint data similar to the voice data, presents a question based on the personal data to identify a user in the air-conditioning area on the basis of an answer acquired from the user in the air-conditioning area.