Files
nexus/wiki/concepts/InclusiveVisuals.md

40 lines
1.8 KiB
Markdown
Raw Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
---
title: "Inclusive Visuals"
type: concept
tags: [generative-ai, bias-mitigation, ai-ethics, image-generation, video-generation]
sources: [design-inclusive-visuals-specialist]
last_updated: 2026-04-24
---
## Definition
AI 生成图像/视频中的包容性视觉呈现——通过结构化提示词工程和负向约束,确保生成内容反映真实多样的社会现实(文化、年龄、残疾、社会经济地位的交叉性),而非刻板印象、符号化表征或文化失真。
## Core Components
### Bias Types to Defeat
- **克隆脸Clone Faces**AI 在生成多样化人群时生成同一人的多个复制版本
- **异域化偏见Exoticism Bias**:对非西方文化的"东方主义"式过度美化或扭曲呈现
- **文化符号乱码Gibberish Cultural Text**:非英语文字、符号生成幻觉
- **地理/建筑失真**:特定文化背景下的环境不符合真实生活现实
- **过度纠正Over-correction**AI 刻意追求多样性时产生符号化、不真实构图
### Technical Interventions
- **显式负向约束**:列举 AI 应避免生成的内容(显式负向提示库)
- **文化真实性锚定**:正确的建筑/服装/布光参数,而非泛化的"多样性"标签
- **交叉性叠加**:同时考虑多重身份维度的叠加表征
- **社区验证**:所描绘社区用户的真实认可
## Related Concepts
- [[NegativePromptingLibrary]]
- [[IntersectionalRepresentation]]
- [[AIArtifactElimination]]
- [[CommunityValidation]]
## Related Agents
- [[InclusiveVisualsSpecialist]](主要执行者)
- [[DesignImagePromptEngineer]](图像提示能力支撑)
- [[UX-Researcher]]QA 验证)
## Related Entities
- [[TheAgency]](所属 Agent 体系)