Human curated dataset of ai generated people images and videos across contexts
HumanCentric-Scenes-AI: A Human-Curated Multimodal Dataset of AI-Generated Human-Centric Scenes Across Surveillance, Occupational Portraits, AI-Edited Images, and Short-Form Video
Computer Vision and Pattern RecognitionArtificial IntelligenceHuman-Computer Interaction
Summary
It is hard to study how people react to AI-generated images and videos of people because good examples are hard to find. The authors created a carefully chosen set of nearly 300 realistic AI-made pictures and short videos showing people in places like surveillance cameras, workplace portraits, and edited real photos. They also provide details like what gender the images show and how hard it is to detect edits in such images. This dataset helps people who want to understand human perception of AI images and improve tools to detect AI-generated content.
What this means in practice
- •For computer vision engineers: Test AI-generated image detection algorithms using a diverse dataset of human-centric scenes in surveillance, occupations, and video formats.
- •For user experience researchers: Assess how people perceive AI-generated images in various real-world contexts to improve design and trust in AI media.
Authors
Nima Kamali, Sana Nasiri, Negar Kamali
Abstract
We introduce HumanCentric-Scenes-AI, a human-curated dataset of 296 AI-generated stimuli that depict people across four contexts in which people encounter AI-generated media. The Surveillance subset contains 85 AI-generated CCTV-style images. The Occupations subset contains 128 AI-generated occupational portraits in a balanced paired design covering 64 occupations with one female and one male version of each. The AI-edited subset contains 27 real photographs in which a region was altered or replaced by a generative model. The Video subset contains 56 AI-generated short videos. We curated the images and videos manually over multiple rounds to make them as photorealistic as possible. For each stimulus we release the file, the generation prompt where recorded, and per-stimulus metadata. We report depicted-gender distributions from an initial set of 134 generations without a specified gender and compare them with U.S. Bureau of Labor Statistics occupational gender shares. We report edit-region size and its association with zero-shot detection difficulty for the AI-edited subset. We report image-quality distributions for the Surveillance subset and subject-category coverage for the Video subset. HumanCentric-Scenes-AI is intended first for human-perception research, and also for calibration of AI-detection classifiers and bias analysis of generative models.