Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

AI-rays: Exploring Bias in the Gaze of AI Through a Multimodal Interactive Installation

Ziyao Gao, Yiwen Zhang, Ling Li, Theodoros Papatheodorou, Wei Zeng

TL;DR

AI-rays is introduced, an interactive installation where AI generates speculative identities from participants’ appearance which are expressed through synthesized personal items placed in participants’ bags, metaphorically highlighting AI’s scrutiny and biases.

Abstract

Data surveillance has become more covert and pervasive with AI algorithms, which can result in biased social classifications. Appearance offers intuitive identity signals, but what does it mean to let AI observe and speculate on them? We introduce AI-rays, an interactive installation where AI generates speculative identities from participants' appearance which are expressed through synthesized personal items placed in participants' bags. It uses speculative X-ray visions to contrast reality with AI-generated assumptions, metaphorically highlighting AI's scrutiny and biases. AI-rays promotes discussions on modern surveillance and the future of human-machine reality through a playful, immersive experience exploring AI biases.

AI-rays: Exploring Bias in the Gaze of AI Through a Multimodal Interactive Installation

TL;DR

AI-rays is introduced, an interactive installation where AI generates speculative identities from participants’ appearance which are expressed through synthesized personal items placed in participants’ bags, metaphorically highlighting AI’s scrutiny and biases.

Abstract

Data surveillance has become more covert and pervasive with AI algorithms, which can result in biased social classifications. Appearance offers intuitive identity signals, but what does it mean to let AI observe and speculate on them? We introduce AI-rays, an interactive installation where AI generates speculative identities from participants' appearance which are expressed through synthesized personal items placed in participants' bags. It uses speculative X-ray visions to contrast reality with AI-generated assumptions, metaphorically highlighting AI's scrutiny and biases. AI-rays promotes discussions on modern surveillance and the future of human-machine reality through a playful, immersive experience exploring AI biases.
Paper Structure (12 sections, 5 figures)

This paper contains 12 sections, 5 figures.

Table of Contents

  1. Inspiration
  2. Concept Design
  3. Identity Narratives of Personal Items
  4. Semantic Transformations of AI-powered X-ray Images
  5. Related Work
  6. The Installation: AI-rays
  7. X-ray Visual Effect Creation
  8. AI-rays Pipeline
  9. Installation Design
  10. Feedback and Discussion
  11. Bias Cases within AI Gaze
  12. Human Perceptions of Machine Scrutiny and Bias

Figures (5)

  • Figure 1: The custom pipeline we constructed to build a LoRA model that can output generative X-ray images.
  • Figure 2: A sample subset from the items dataset used to augment participants’ images.
  • Figure 3: The AI-rays system pipeline accepts a participant’s photo as input (left) and outputs the generative X-ray image along with hypothetical personal items (right). To achieve this it uses three core modules we built that understand, composite, and generate images (center).
  • Figure 4: Generated X-ray images with the speculated personal items. These examples show that women are associated with beauty products and less often with items linked to professions. Additionally, a man in Arab robes is perceived as having a better economic status compared to a man in casual attire.
  • Figure 5: Examples of bias from wider tests based on three representative dimensions: skin color, gender, and occupations using the Inference Module from our pipeline.