Related Experiment Video
Updated: Jan 12, 2026

Central and Divided Visual Field Presentation of Emotional Images to Measure Hemispheric Differences in Motivated Attention
Published on: November 16, 2017
A Prompt Engineering Method for Generating Emotional Images for Psychological Research
Yuqing Lu1,2, Xin Hu3, Dan Zhang1,2
1Department of Psychological and Cognitive Sciences, Tsinghua University, Beijing, China.
Artificial Intelligence (AI) image generation offers a scalable method for creating emotional image datasets for psychological research. This AI-generated emotional image set (AGEIS) effectively elicits target emotions and appears authentic.
Area of Science:
- Psychology
- Computer Science
- Artificial Intelligence
Background:
- Standardized image datasets are crucial for emotion research.
- Traditional methods for dataset creation are labor-intensive and limited in scope.
- AI-driven Text-to-Image Synthesis (TIS) presents a potential solution for generating emotional images.
Purpose of the Study:
- To introduce and evaluate a novel AI-based procedure for generating emotional images for psychological research.
- To create an AI-Generated Emotional Image Set (AGEIS) covering seven distinct emotion categories.
- To assess the effectiveness and authenticity of AI-generated emotional images.
Main Methods:
- Developed a novel procedure involving Seed Prompt Formation, Prompt Set Generation, and TIS.
- Utilized AI tools like Midjourney and ChatGPT for image generation.
- Conducted human evaluations to assess emotion induction and perceived authenticity of generated images.
Main Results:
- The AI-generated images in AGEIS effectively and specifically induced target emotions.
- AGEIS images were perceived as comparably authentic to real photos.
- Images depicting disgust and tenderness were more frequently identified as AI-generated.
Conclusions:
- The proposed AI procedure offers a cost-effective and scalable method for generating emotional images.
- This approach can facilitate large-scale and longitudinal studies in emotion research.
- AI-generated images hold significant potential for advancing psychological research methodologies.
Related Concept Videos
Facial Feedback Hypothesis
Cognitive Theories: Schachter-Singer Theory of Emotion
Physiological Arousal and Cognitive Labeling
According to this theory, when an individual experiences...
Physiological Theories: James-Lange Theory of Emotion
Labeling Emotion
Emotional Expression
Universal Facial Expressions
Psychologist Paul Ekman identified seven basic...

