Identifying Implicit Social Biases in Vision-Language Models

• URL: http://arxiv.org/abs/2411.00997v1
• Date: Fri, 01 Nov 2024 19:41:28 GMT
• Title: Identifying Implicit Social Biases in Vision-Language Models
• Authors: Kimia Hamidieh, Haoran Zhang, Walter Gerych, Thomas Hartvigsen, Marzyeh Ghassemi,
• Abstract summary: We conduct a systematic analysis of the social biases that are present in vision-language models. We find that CLIP frequently displays undesirable associations between harmful words and specific demographic groups. Our findings highlight the importance of evaluating and addressing bias in vision-language models.
• Score: 34.53206726136747
• License:
• Abstract: Vision-language models, like CLIP (Contrastive Language Image Pretraining), are becoming increasingly popular for a wide range of multimodal retrieval tasks. However, prior work has shown that large language and deep vision models can learn historical biases contained in their training sets, leading to perpetuation of stereotypes and potential downstream harm. In this work, we conduct a systematic analysis of the social biases that are present in CLIP, with a focus on the interaction between image and text modalities. We first propose a taxonomy of social biases called So-B-IT, which contains 374 words categorized across ten types of bias. Each type can lead to societal harm if associated with a particular demographic group. Using this taxonomy, we examine images retrieved by CLIP from a facial image dataset using each word as part of a prompt. We find that CLIP frequently displays undesirable associations between harmful words and specific demographic groups, such as retrieving mostly pictures of Middle Eastern men when asked to retrieve images of a "terrorist". Finally, we conduct an analysis of the source of such biases, by showing that the same harmful stereotypes are also present in a large image-text dataset used to train CLIP models for examples of biases that we find. Our findings highlight the importance of evaluating and addressing bias in vision-language models, and suggest the need for transparency and fairness-aware curation of large pre-training datasets.

Related papers

• Spoken Stereoset: On Evaluating Social Bias Toward Speaker in Speech Large Language Models [50.40276881893513]
  This study introduces Spoken Stereoset, a dataset specifically designed to evaluate social biases in Speech Large Language Models (SLLMs) By examining how different models respond to speech from diverse demographic groups, we aim to identify these biases. The findings indicate that while most models show minimal bias, some still exhibit slightly stereotypical or anti-stereotypical tendencies.
  arXiv  Detail & Related papers  (2024-08-14T16:55:06Z)
• Dataset Scale and Societal Consistency Mediate Facial Impression Bias in Vision-Language AI [17.101569078791492]
  We study 43 CLIP vision-language models to determine whether they learn human-like facial impression biases. We show for the first time that the the degree to which a bias is shared across a society predicts the degree to which it is reflected in a CLIP model.
  arXiv  Detail & Related papers  (2024-08-04T08:26:58Z)
• Blind Dates: Examining the Expression of Temporality in Historical Photographs [57.07335632641355]
  We investigate the dating of images using OpenCLIP, an open-source implementation of CLIP, a multi-modal language and vision model. We use the textitDe Boer Scene Detection dataset, containing 39,866 gray-scale historical press photographs from 1950 to 1999. Our analysis reveals that images featuring buses, cars, cats, dogs, and people are more accurately dated, suggesting the presence of temporal markers.
  arXiv  Detail & Related papers  (2023-10-10T13:51:24Z)
• Vocabulary-free Image Classification [75.38039557783414]
  We formalize a novel task, termed as Vocabulary-free Image Classification (VIC) VIC aims to assign to an input image a class that resides in an unconstrained language-induced semantic space, without the prerequisite of a known vocabulary. CaSED is a method that exploits a pre-trained vision-language model and an external vision-language database to address VIC in a training-free manner.
  arXiv  Detail & Related papers  (2023-06-01T17:19:43Z)
• Uncurated Image-Text Datasets: Shedding Light on Demographic Bias [21.421722941901123]
  Even small but manually annotated datasets, such as MSCOCO, are affected by societal bias. Our first contribution is to annotate part of the Google Conceptual Captions dataset, widely used for training vision-and-language models. Second contribution is to conduct a comprehensive analysis of the annotations, focusing on how different demographic groups are represented. Third contribution is to evaluate three prevailing vision-and-language tasks, showing that societal bias is a persistent problem in all of them.
  arXiv  Detail & Related papers  (2023-04-06T02:33:51Z)
• DeAR: Debiasing Vision-Language Models with Additive Residuals [5.672132510411465]
  Large pre-trained vision-language models (VLMs) provide rich, adaptable image and text representations. These models suffer from societal biases owing to the skewed distribution of various identity groups in the training data. We present DeAR, a novel debiasing method that learns additive residual image representations to offset the original representations.
  arXiv  Detail & Related papers  (2023-03-18T14:57:43Z)
• Discovering and Mitigating Visual Biases through Keyword Explanation [66.71792624377069]
  We propose the Bias-to-Text (B2T) framework, which interprets visual biases as keywords. B2T can identify known biases, such as gender bias in CelebA, background bias in Waterbirds, and distribution shifts in ImageNet-R/C. B2T uncovers novel biases in larger datasets, such as Dollar Street and ImageNet.
  arXiv  Detail & Related papers  (2023-01-26T13:58:46Z)
• Worst of Both Worlds: Biases Compound in Pre-trained Vision-and-Language Models [17.90351661475405]
  This work extends text-based bias analysis methods to investigate multimodal language models. We demonstrate that VL-BERT exhibits gender biases, often preferring to reinforce a stereotype over faithfully describing the visual scene.
  arXiv  Detail & Related papers  (2021-04-18T00:02:32Z)
• Probing Contextual Language Models for Common Ground with Visual Representations [76.05769268286038]
  We design a probing model that evaluates how effective are text-only representations in distinguishing between matching and non-matching visual representations. Our findings show that language representations alone provide a strong signal for retrieving image patches from the correct object categories. Visually grounded language models slightly outperform text-only language models in instance retrieval, but greatly under-perform humans.
  arXiv  Detail & Related papers  (2020-05-01T21:28:28Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.