Multi-part touch emotion recognition method and system of emotion accompanying robot

Abstract

The invention discloses a multi-part touch emotion recognition method and system of an emotion accompanying robot, and belongs to the technical field of intelligent robots. The method comprises the steps that touch signals are collected through sensor arrays arranged on the head, the chest, the arms and other parts of the robot; processing the signal and extracting multi-dimensional features such as pressure grade, duration, frequency and part time difference; and matching the feature vector with a pre-stored emotion rule table, solving conflicts by adopting a priority strategy, and finally outputting an identified emotion type to drive interactive feedback. According to the method and the device, the problems of low single-point touch emotion recognition precision and unnatural interaction in the prior art are solved, accurate and reliable recognition of various fine and smooth emotions can be completed only through a tactile channel, and the man-machine emotion interaction experience is remarkably improved.

Description

Technical Field

[0001] This invention belongs to the field of intelligent robot technology, and in particular relates to a multi-part touch emotion recognition method and system for emotional companion robots, which is used to improve the robot's ability to perceive the user's emotional state and the naturalness of the interaction. Background Technology

[0002] With the development of artificial intelligence technology, emotional companion robots are gradually becoming important tools for providing emotional support and interactive companionship. Currently, emotional robots mostly rely on speech recognition, computer vision (such as facial expression recognition), or posture analysis to determine the user's emotions. However, these methods have significant limitations: their recognition efficiency drops drastically in poor lighting, noisy environments, or when the user's back is to the robot. More importantly, they cannot effectively capture the emotional information conveyed by the user through physical contact (such as touching, patting, or hugging).

[0003] Currently, while some research attempts to integrate touch sensors into robots to detect emotions, most rely on single, isolated touch points. This approach fails to consider the different meanings of various body parts (such as the head, arms, and chest) in conveying emotions and ignores the dynamic characteristics of touch (such as pressure changes, duration, and frequency sequences). For example, a brief tap on the head conveys a completely different emotion than a long, gentle pat on the head, and a single sensor cannot distinguish this difference. This results in low accuracy and a high false positive rate in current emotion recognition technologies, leading to stiff and unnatural robot interaction feedback, severely impacting the user experience.

[0004] Therefore, there is an urgent need in this field for a technical solution that can comprehensively utilize touch information from multiple parts and their dynamic characteristics to achieve more accurate and natural emotion recognition. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to provide a method, system and emotional companion robot for multi-part touch emotion recognition.

[0006] This invention is implemented as follows:

[0007] In a first aspect, the present invention provides a method for multi-part touch emotion recognition in an emotional companion robot, comprising the following steps:

[0008] S1: By deploying touch sensors at at least two different parts of the robot, raw signals generated by user touch are collected, including pressure data, position data and time data;

[0009] S2: Digitize and process the original signal to extract a multi-dimensional feature vector for emotion determination. The multi-dimensional feature vector includes pressure level, touch duration, touch frequency, and time difference between touches on different parts of the body.

[0010] S3: Match the multidimensional feature vector with a pre-stored emotion recognition rule table, which defines the mapping relationship between various emotion types and feature vectors;

[0011] S4: Output the successfully matched emotion type to drive the robot to make corresponding emotional responses.

[0012] Furthermore, in step S3, when the multidimensional feature vector simultaneously satisfies the recognition rules for multiple emotion types, a decision is made according to a preset priority order, and the emotion type with the highest priority is output; the priority order is set based on the urgency and intensity of the emotion.

[0013] Furthermore, the priority order is as follows: anger > anxiety > dissatisfaction > frustration > excitement > warmth > intimacy > encouragement > care > comfort > pleasure > attention.

[0014] Furthermore, the pressure level is quantified into three levels: "light touch", "medium touch", and "heavy touch".

[0015] Furthermore, the mapping relationships defined in the emotion recognition rule table include:

[0016] Anger: Triggered by a heavy touch on any part of the body, or a rapid touch with a frequency of 4Hz or higher lasting ≥1 second;

[0017] Anxiety is triggered by irregular light touches on multiple body parts, occurring ≥ 5 times within 1 second with irregular intervals.

[0018] Intimacy: Triggered by a light or medium touch to the head lasting ≥3 seconds;

[0019] Warm Emotion: Triggered by a synchronized light or medium touch of both arms, with a time difference of ≤0.5 seconds and a duration of ≥1 second.

[0020] Secondly, the present invention provides a multi-part touch emotion recognition system for an emotional companion robot, used to implement the above-mentioned multi-part touch emotion recognition method for an emotional companion robot, the emotion recognition system comprising:

[0021] The signal acquisition module is used to acquire raw analog signals through a multi-touch sensor and then digitize them.

[0022] A data processing module, connected to the signal acquisition module, is used to extract the multidimensional feature vector;

[0023] The emotion determination module is connected to the data processing module. It has pre-stored the emotion recognition rule table and priority order, and is used to perform emotion matching and decision-making.

[0024] The control system, connected to the emotion determination module, is used to control the robot to perform feedback actions based on the emotion result.

[0025] Furthermore, the touch sensors are deployed on the robot's head, chest, left arm, and right arm.

[0026] Furthermore, the sampling frequency of the signal acquisition module is not less than 10Hz.

[0027] Thirdly, the present invention provides an emotional companion robot that integrates a multi-part touch emotion recognition system as described above.

[0028] The advantages of this invention are:

[0029] 1. Multi-dimensional perception: By working together with sensors in multiple locations, the system comprehensively analyzes multi-dimensional features such as pressure, time, and location combinations, greatly enriching the dimensions of emotional information collection and overcoming the limitations of single-point touch perception.

[0030] 2. High recognition accuracy and high reliability: The recognition rules are formulated based on clear hardware signal characteristics. The algorithm logic is clear and the amount of computation is small. It is very suitable for stable operation in resource-constrained embedded systems, ensuring high real-time performance and high reliability.

[0031] 3. Enhanced naturalness of interaction: It can accurately identify up to 12 subtle emotions using only the tactile channel, making the robot's feedback more emotionally targeted and layered, greatly improving the naturalness of human-computer interaction and user immersion.

[0032] 4. Strong anti-interference capability: It does not rely on visual and auditory sensors that are easily affected by the environment, and can still provide a stable emotional interaction experience in special environments such as darkness and noise. Attached Figure Description

[0033] The present invention will now be further described with reference to the accompanying drawings and embodiments.

[0034] Figure 1 This is a flowchart of the overall workflow of the multi-part touch emotion recognition system of the present invention.

[0035] Figure 2 This is a schematic diagram of the emotion recognition rule table and priority used in the embodiments of the present invention. Detailed Implementation

[0036] This invention provides a method, system, and robot for multi-part touch emotion recognition in an emotional companion robot. By deploying multi-point touch sensors, setting differentiated weights, and introducing dynamic feature analysis, it solves the problem of inaccurate single-point touch perception, thereby improving the accuracy of emotion recognition and the naturalness of interaction.

[0037] Example 1:

[0038] like Figure 1 As shown, this embodiment provides a multi-part touch emotion recognition method for an emotional companion robot, including the following steps:

[0039] S1: The raw signals generated by the user's touch are collected by touch sensors deployed at least two different parts of the robot. The raw signals include pressure data, position data and time data.

[0040] S2: The original signal is digitized and processed to extract a multi-dimensional feature vector for emotion determination. The multi-dimensional feature vector includes pressure level, touch duration, touch frequency and time difference between touches at different parts; the pressure level is quantified into three levels: "light touch", "medium touch" and "heavy touch".

[0041] S3: Match the multidimensional feature vector with a pre-stored emotion recognition rule table, which defines the mapping relationship between various emotion types and feature vectors;

[0042] When a multidimensional feature vector simultaneously satisfies the recognition rules for multiple emotion types, a decision is made based on a preset priority order, and the emotion type with the highest priority is output; the priority order is set based on the urgency and intensity of the emotion.

[0043] The priority order is: anger > anxiety > dissatisfaction > frustration > excitement > warmth > intimacy > encouragement > care > comfort > pleasure > attention.

[0044] The mapping relationships defined in the emotion recognition rule table include:

[0045] Anger: Triggered by a heavy touch on any part of the body, or a rapid touch with a frequency of 4Hz or higher lasting ≥1 second;

[0046] Anxiety is triggered by irregular light touches on multiple body parts, occurring ≥ 5 times within 1 second with irregular intervals.

[0047] Intimacy: Triggered by a light or medium touch to the head lasting ≥3 seconds;

[0048] Warm Emotion: Triggered by a synchronized light or medium touch of both arms, with a time difference of ≤0.5 seconds and a duration of ≥1 second.

[0049] Figure 2The diagram shows the emotion recognition rule table and its priorities.

[0050] S4: Output the successfully matched emotion type to drive the robot to make corresponding emotional responses.

[0051] Example 2:

[0052] This embodiment provides a multi-part touch emotion recognition system for an emotional companion robot, used to implement the multi-part touch emotion recognition method for an emotional companion robot described in Embodiment 2. The emotion recognition system includes:

[0053] The signal acquisition module is used to acquire raw analog signals and digitize them through multi-point touch sensors; the touch sensors are deployed on the robot's head, chest, left arm, and right arm;

[0054] A data processing module, connected to the signal acquisition module, is used to extract the multidimensional feature vector; the sampling frequency of the signal acquisition module is not less than 10Hz;

[0055] The emotion determination module is connected to the data processing module. It has pre-stored the emotion recognition rule table and priority order, and is used to perform emotion matching and decision-making.

[0056] The control system, connected to the emotion determination module, is used to control the robot to perform feedback actions based on the emotion result.

[0057] Example 3:

[0058] This embodiment provides an emotional companion robot, which integrates a multi-part touch emotion recognition system for an emotional companion robot as described in Embodiment 2.

[0059] The above embodiments and figures are not intended to limit the product form and style of the present invention. Any appropriate changes or modifications made by those skilled in the art should be considered as not departing from the patent scope of the present invention.

Claims

1. A method for multi-part touch emotion recognition in an emotional companion robot, characterized in that: Includes the following steps: S1: By deploying touch sensors at at least two different parts of the robot, raw signals generated by user touch are collected, including pressure data, position data and time data; S2: Digitize and process the original signal to extract a multi-dimensional feature vector for emotion determination. The multi-dimensional feature vector includes pressure level, touch duration, touch frequency, and time difference between touches on different parts of the body. S3: Match the multidimensional feature vector with a pre-stored emotion recognition rule table, which defines the mapping relationship between various emotion types and feature vectors; S4: Output the successfully matched emotion type to drive the robot to make corresponding emotional responses.

2. The multi-part touch emotion recognition method for an emotional companion robot according to claim 1, characterized in that: In step S3, when the multidimensional feature vector simultaneously satisfies the recognition rules for multiple emotion types, a decision is made according to a preset priority order, and the emotion type with the highest priority is output; the priority order is set based on the urgency and intensity of the emotion.

3. The multi-part touch emotion recognition method for an emotional companion robot according to claim 2, characterized in that: The priority order is: anger > anxiety > dissatisfaction > frustration > excitement > warmth > intimacy > encouragement > care > comfort > pleasure > attention.

4. The multi-part touch emotion recognition method for an emotional companion robot according to claim 1, characterized in that: The pressure levels are quantified into three levels: "light touch", "medium touch", and "heavy touch".

5. The multi-part touch emotion recognition method for an emotional companion robot according to claim 1, characterized in that: The mapping relationships defined in the emotion recognition rule table include: Anger: Triggered by a heavy touch on any part of the body, or a rapid touch with a frequency of 4Hz or higher lasting ≥1 second; Anxiety is triggered by irregular light touches on multiple body parts, occurring ≥ 5 times within 1 second with irregular intervals. Intimacy: Triggered by a light or medium touch to the head lasting ≥3 seconds; Warm Emotion: Triggered by a synchronized light or medium touch of both arms, with a time difference of ≤0.5 seconds and a duration of ≥1 second.

6. A multi-part touch emotion recognition system for an emotional companion robot, characterized in that: A method for multi-part touch emotion recognition in an emotional companion robot according to any one of claims 1-5, the emotion recognition system comprising: The signal acquisition module is used to acquire raw analog signals through a multi-touch sensor and then digitize them. A data processing module, connected to the signal acquisition module, is used to extract the multidimensional feature vector; The emotion determination module is connected to the data processing module. It has pre-stored the emotion recognition rule table and priority order, and is used to perform emotion matching and decision-making. The control system, connected to the emotion determination module, is used to control the robot to perform feedback actions based on the emotion result.

7. The multi-part touch emotion recognition system for an emotional companion robot according to claim 6, characterized in that: The touch sensors are deployed on the robot's head, chest, left arm, and right arm.

8. The multi-part touch emotion recognition system for an emotional companion robot according to claim 6, characterized in that: The sampling frequency of the signal acquisition module is not less than 10Hz.

9. An emotional companion robot, characterized in that: The system integrates a multi-part touch emotion recognition system for an emotional companion robot as described in any one of claims 6-8.

Images