Stereotyping and Bias in the Flickr30K Dataset

2016/05/19 by Emiel van Miltenburg, van Miltenburg, Emiel · 2 citations
Biochemistry, Genetics and Molecular Biology · Medicine · #Bioinformatics and Genomic Networks #Cell Image Analysis Techniques #Computation and Language (cs.CL) #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #Radiomics and Machine Learning in Medical Imaging

paper · pdf · doi:10.48550/arxiv.1605.06083

openalex publication_date 2016/05/19 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

An untested assumption behind the crowdsourced descriptions of the images in the Flickr30K dataset (Young et al., 2014) is that they "focus only on the information that can be obtained from the image alone" (Hodosh et al., 2013, p. 859). This paper presents some evidence against this assumption, and provides a list of biases and unwarranted inferences that can be found in the Flickr30K dataset. Finally, it considers methods to find examples of these, and discusses how we should deal with stereotype-driven descriptions in future applications.

Cited by

• ‘We can see a savage’: a case study of the colonial gaze in generative AI algorithms
• Understanding Gender Bias in AI-Generated Product Descriptions

Related

• Uncurated Image-Text Datasets: Shedding Light on Demographic Bias
• A Friendly Face: Do Text-to-Image Systems Rely on Stereotypes when the Input is Under-Specified?
• Resampled Datasets Are Not Enough: Mitigating Societal Bias Beyond Single Attributes
• Seeing through the Human Reporting Bias: Visual Classifiers from Noisy\n Human-Centric Labels
• BBQ-V: Benchmarking Visual Stereotype Bias in Large Multimodal Models