Emotion recognition and analysis system for learning pictures

Abstract

The invention relates to the technical field of computer vision and artificial intelligence, and particularly discloses an emotion recognition and analysis system for learning pictures. The system comprises a modal data acquisition module, a feature automatic learning module, an emotion classification and recognition module, an emotion intensity analysis module, a feedback and optimization mechanism and a visualization and interaction interface. The system can adapt to complex and changeable emotion expression scenes.

Description

Technical Field

[0001] The present invention relates to the field of computer vision and artificial intelligence technology, and in particular to an emotion recognition and analysis system for learning pictures. Background Art

[0002] With the rapid development of artificial intelligence technology, emotion recognition technology has shown great potential in the fields of human-computer interaction, intelligent monitoring, and mental health monitoring. Traditional emotion recognition methods rely on manually designed feature extraction and classification models, which are not only time-consuming and labor-intensive, but also difficult to adapt to complex and changing emotional expression scenarios. Therefore, it is particularly important to develop a system that can automatically learn and accurately identify the emotions of people in images or videos. Summary of the invention

[0003] The present invention provides an emotion recognition and analysis system for learning pictures, which can overcome certain defects of the prior art.

[0004] According to the present invention, a learning picture emotion recognition and analysis system includes: a modal data acquisition module, a feature automatic learning module, an emotion classification and recognition module, an emotion intensity analysis module, a feedback and optimization mechanism, and a visualization and interactive interface.

[0005] Through the present invention, the modal data acquisition module can collect data of multiple modalities in real time and accurately, such as images, audio, video, etc., to provide a rich information basis for subsequent sentiment analysis; it supports multiple sensors and data sources, and improves the flexibility and adaptability of the system; The feature automatic learning module uses advanced technologies such as deep learning to automatically extract emotion-related features from raw data, eliminating the need to manually design feature extraction rules and reducing development costs. The emotion classification and recognition module can map the extracted features to emotion labels to achieve automatic emotion classification and recognition; The sentiment intensity analysis module helps to more accurately assess the user's emotional state and provide a basis for subsequent decision-making and optimization; The feedback and optimization mechanism can automatically adjust model parameters based on user feedback or new data input to optimize the accuracy and generalization ability of emotion recognition; The visualization and interactive interface provides an intuitive and easy-to-use user interface, allowing users to easily query and annotate data, and adjust system configuration.

[0006] Preferably, the multimodal data acquisition module is capable of synchronously acquiring facial images, body movement data and audio data, and performing preprocessing to form a standardized multimodal data set.

[0007] Through the present invention, data in multiple modalities, such as facial images, body movements and audio, are collected synchronously, so that the user's emotional expression can be captured and analyzed in an all-round way; data in multiple modalities complement each other, which can more accurately reflect the user's true emotional state and improve the accuracy of emotion recognition.

[0008] Preferably, the automatic feature learning module adopts a deep learning network to automatically extract emotion-related features from multimodal data.

[0009] Through the present invention, the deep learning network has a powerful feature learning ability and can automatically extract features closely related to emotions from the original data without relying on manually designed feature extraction methods; it avoids the subjectivity and incompleteness of manual feature extraction and improves the accuracy and efficiency of feature extraction.

[0010] Preferably, the emotion classification and recognition module is capable of recognizing at least six basic emotion categories, and optionally recognizing more complex emotion states.

[0011] Through the present invention, the recognition of six basic emotion categories has covered the common emotional states in people's daily lives, and the ability to recognize more complex emotional states further expands the recognition scope of the system; this enables the system to more comprehensively understand and analyze the user's emotional state, providing richer information for subsequent decision-making and optimization.

[0012] Preferably, the emotion intensity analysis module is capable of quantitatively analyzing the intensity or degree of emotion.

[0013] Through the present invention, the emotion intensity analysis module can not only identify the category of emotion, but also further quantify the intensity or degree of emotion, thereby providing more accurate and detailed emotion information; it helps to understand the user's emotional state more deeply and provide a more accurate basis for subsequent decision-making or feedback.

[0014] Preferably, the feedback and optimization mechanism automatically adjusts model parameters based on user feedback or new data input.

[0015] Through the present invention, by continuously receiving user feedback and new data input, the system can evaluate the performance of the current model in real time and automatically adjust the model parameters accordingly; this helps the system to continuously learn and improve, thereby continuously improving the accuracy and robustness of the model.

[0016] Preferably, the visualization and interaction interface provides a user-friendly interface design and supports querying recognition results, annotated data, and adjustment of system configuration.

[0017] Through the present invention, the user-friendly interface design enables users to easily get started and quickly become familiar with system operations without complicated training or guidance; it supports functions such as querying recognition results and annotating data, so that users can intuitively understand the working status and output results of the system, which is convenient for subsequent analysis and processing; through the visualization and interactive interface, users can easily participate in the data annotation and configuration adjustment of the system, so as to interact with the system more actively; this participation helps to enhance users' trust and satisfaction with the system and promote good cooperation between users and the system. DETAILED DESCRIPTION

[0018] In order to further understand the content of the present invention, the present invention is described in detail in conjunction with the embodiments. It should be understood that the embodiments are only for explaining the present invention and are not intended to limit it.

[0019] Example 1 This embodiment provides an emotion recognition and analysis system for learning pictures, including a modal data acquisition module, a feature automatic learning module, an emotion classification and recognition module, an emotion intensity analysis module, a feedback and optimization mechanism, and a visualization and interactive interface.

[0020] Through this embodiment, the modal data acquisition module can collect data of multiple modalities in real time and accurately, such as images, audio, video, etc., to provide a rich information basis for subsequent sentiment analysis; it supports multiple sensors and data sources, and improves the flexibility and adaptability of the system; The feature automatic learning module uses advanced technologies such as deep learning to automatically extract emotion-related features from raw data, eliminating the need to manually design feature extraction rules and reducing development costs. The emotion classification and recognition module can map the extracted features to emotion labels to achieve automatic emotion classification and recognition; The sentiment intensity analysis module helps to more accurately assess the user's emotional state and provide a basis for subsequent decision-making and optimization; The feedback and optimization mechanism can automatically adjust model parameters based on user feedback or new data input to optimize the accuracy and generalization ability of emotion recognition; The visualization and interactive interface provides an intuitive and easy-to-use user interface, allowing users to easily query and annotate data, and adjust system configuration.

[0021] In this embodiment, the multimodal data acquisition module can synchronously acquire facial images, body movement data, and audio data, and perform preprocessing to form a standardized multimodal data set.

[0022] Through this embodiment, by synchronously collecting data in multiple modalities, such as facial images, body movements and audio, the user's emotional expression can be captured and analyzed in an all-round way; the data in multiple modalities complement each other, and can more accurately reflect the user's true emotional state, thereby improving the accuracy of emotion recognition.

[0023] In this embodiment, the automatic feature learning module uses a deep learning network to automatically extract emotion-related features from multimodal data.

[0024] Through this embodiment, the deep learning network has a powerful feature learning ability and can automatically extract features closely related to emotions from the original data without relying on manually designed feature extraction methods; it avoids the subjectivity and incompleteness of manual feature extraction and improves the accuracy and efficiency of feature extraction.

[0025] In this embodiment, the emotion classification and recognition module is capable of recognizing at least six basic emotion categories, and optionally recognizing more complex emotion states.

[0026] Through this embodiment, the recognition of six basic emotion categories has covered the common emotional states in people's daily lives, and the ability to recognize more complex emotional states further expands the recognition scope of the system; this enables the system to more comprehensively understand and analyze the user's emotional state, providing richer information for subsequent decision-making and optimization.

[0027] In this embodiment, the emotion intensity analysis module can quantitatively analyze the intensity or degree of emotion.

[0028] Through this embodiment, the emotion intensity analysis module can not only identify the category of emotion, but also further quantify the intensity or degree of emotion, thereby providing more accurate and detailed emotion information; it helps to understand the user's emotional state more deeply and provide a more accurate basis for subsequent decision-making or feedback.

[0029] In this embodiment, the feedback and optimization mechanism automatically adjusts the model parameters according to user feedback or new data input.

[0030] Through this embodiment, by continuously receiving user feedback and new data input, the system can evaluate the performance of the current model in real time and automatically adjust the model parameters accordingly; this helps the system to continuously learn and improve, thereby continuously improving the accuracy and robustness of the model.

[0031] In this embodiment, the visualization and interaction interface provides a user-friendly interface design and supports querying recognition results, annotated data, and adjustment of system configuration.

[0032] Through this embodiment, the user-friendly interface design enables users to easily get started and quickly become familiar with system operations without complicated training or guidance; it supports functions such as querying recognition results and annotating data, so that users can intuitively understand the system's working status and output results, facilitating subsequent analysis and processing; through the visualization and interactive interface, users can easily participate in the system's data annotation and configuration adjustments, thereby interacting with the system more actively; this level of participation helps to enhance users' trust and satisfaction with the system, and promotes good cooperation between users and the system.

[0033] It is easy to understand that those skilled in the art can combine, split, reorganize, etc. the embodiments of the present application to obtain other embodiments based on one or several embodiments provided in the present application, and these embodiments do not exceed the protection scope of the present application.

[0034] The present invention and its implementation methods are described schematically above, and the description is not restrictive. The embodiments shown in the embodiments are only part of the implementation methods of the present invention, and the actual structure is not limited thereto. Therefore, if a person skilled in the art is inspired by the embodiments and designs a structure and an implementation method similar to the technical solution without creativity without departing from the purpose of the invention, they should all fall within the protection scope of the present invention.

Claims

1. A learning picture emotion recognition and analysis system, characterized by: It includes multimodal data acquisition module, feature automatic learning module, sentiment classification and recognition module, sentiment intensity analysis module, feedback and optimization mechanism, and visualization and interactive interface.

2. The emotion recognition and analysis system for learning pictures according to claim 1, characterized in that: The multimodal data acquisition module can synchronously acquire facial images, body movement data and audio data, and perform preprocessing to form a standardized multimodal data set.

3. The emotion recognition and analysis system for learning pictures according to claim 1, characterized in that: The feature automatic learning module adopts a deep learning network to automatically extract emotion-related features from multimodal data.

4. The system for learning emotion recognition and analysis of pictures according to claim 1, characterized in that: The emotion classification and recognition module is capable of recognizing at least six basic emotion categories and optionally more complex emotion states.

5. The emotion recognition and analysis system for learning pictures according to claim 1, characterized in that: The emotion intensity analysis module can quantitatively analyze the intensity or degree of emotion.

6. The emotion recognition and analysis system for learning pictures according to claim 1, characterized in that: The feedback and optimization mechanism automatically adjusts model parameters based on user feedback or new data input.

7. The emotion recognition and analysis system for learning pictures according to claim 1, characterized in that: The visualization and interaction interface provides a user-friendly interface design and supports querying recognition results, annotating data, and adjusting system configuration.