Social Robot Mood Recognition for Multi-User Feedback
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current robots lack the ability to provide personalized feedback to multiple users based on their emotional states during content output, failing to effectively determine and respond to the mood of a group.
Innovation Solution
A robot equipped with a camera, processor, and output units like displays and speakers, which uses facial expression recognition through deep learning to determine the mood of a group and provides feedback such as facial expressions, voice outputs, or movements based on the recognized emotions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If the robot monitors user state using camera and microphone during content output, then the robot can gather information about user emotions, but the robot cannot provide personalized feedback to multiple users based on their emotional states
Solution Approach 1:
The patent segments the group of multiple users into individual user entities, each with separate emotion recognition and feedback generation. The processor extracts individual face images from the group image, recognizes emotions for each user separately, and generates personalized feedback images for each user based on their specific emotional state, rather than providing a single generic feedback for the entire group.
Solution Approach 2:
The patent introduces an emotion recognition model as an intermediary component between the image acquisition and feedback generation processes. This model analyzes facial expressions to determine emotional states, serving as a mediator that transforms raw image data into meaningful emotion information, which then guides the selection of appropriate personalized feedback images for each user.
2Measurement precision
If the robot provides feedback based on group mood, then user engagement can be enhanced, but the robot cannot effectively determine and respond to individual emotional states within the group
Solution Approach 1:
The patent segments the feedback generation process into individual user streams, where each user's emotion is recognized separately and their personalized feedback is generated independently. This segmentation enables precise emotion measurement for each user while maintaining processing efficiency through parallel化处理 of multiple user feedbacks.
Solution Approach 2:
The patent performs preliminary emotion recognition on individual users before generating feedback images. By first identifying each user's emotional state from the group image and then selecting appropriate feedback images based on these pre-determined emotions, the system achieves both high measurement precision and efficient feedback delivery.
Data Source
AI summary
A robot includes an output unit including at least one of a display or a speaker, a camera, and a processor configured to control the output unit to output content, to acquire an image including a plurality of users through the camera while the content is output, to determine a mood of a group including the plurality of users based on the acquired image, and to control the output unit to output feedback based on the determined mood.