A Prompt Array Keeps the Bias Away: Debiasing Vision-Language Models with Adversarial Learning

• URL: http://arxiv.org/abs/2203.11933v4
• Date: Wed, 26 Oct 2022 03:19:13 GMT
• Title: A Prompt Array Keeps the Bias Away: Debiasing Vision-Language Models with Adversarial Learning
• Authors: Hugo Berg, Siobhan Mackenzie Hall, Yash Bhalgat, Wonsuk Yang, Hannah Rose Kirk, Aleksandar Shtedritski, Max Bain
• Abstract summary: Vision-language models can encode societal biases and stereotypes. There are challenges to measuring and mitigating these multimodal harms. We investigate bias measures and apply ranking metrics for image-text representations.
• Score: 55.96577490779591
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Vision-language models can encode societal biases and stereotypes, but there are challenges to measuring and mitigating these multimodal harms due to lacking measurement robustness and feature degradation. To address these challenges, we investigate bias measures and apply ranking metrics for image-text representations. We then investigate debiasing methods and show that prepending learned embeddings to text queries that are jointly trained with adversarial debiasing and a contrastive loss reduces various bias measures with minimal degradation to the image-text representation.

Related papers

• When Text and Images Don't Mix: Bias-Correcting Language-Image Similarity Scores for Anomaly Detection [35.09035417676343]
  We show that the embeddings of text inputs unexpectedly tightly cluster together, far away from image embeddings, contrary to the model's contrastive training objective. We propose a novel methodology called BLISS which directly accounts for this similarity bias through the use of an auxiliary, external set of text inputs.
  arXiv  Detail & Related papers  (2024-07-24T08:20:02Z)
• Information Theoretic Text-to-Image Alignment [49.396917351264655]
  We present a novel method that relies on an information-theoretic alignment measure to steer image generation. Our method is on-par or superior to the state-of-the-art, yet requires nothing but a pre-trained denoising network to estimate MI.
  arXiv  Detail & Related papers  (2024-05-31T12:20:02Z)
• Exposing Text-Image Inconsistency Using Diffusion Models [36.820267498751626]
  A growing problem is text-image inconsistency, where images are misleadingly paired with texts with different intent or meaning. This study introduces D-TIIL, which employs text-to-image diffusion models to localize semantic inconsistencies in text and image pairs. D-TIIL offers a scalable and evidence-based approach to identifying and localizing text-image inconsistency.
  arXiv  Detail & Related papers  (2024-04-28T00:29:24Z)
• Few-Shot Adversarial Prompt Learning on Vision-Language Models [62.50622628004134]
  The vulnerability of deep neural networks to imperceptible adversarial perturbations has attracted widespread attention. Previous efforts achieved zero-shot adversarial robustness by aligning adversarial visual features with text supervision. We propose a few-shot adversarial prompt framework where adapting input sequences with limited data makes significant adversarial robustness improvement.
  arXiv  Detail & Related papers  (2024-03-21T18:28:43Z)
• Debiasing Multimodal Large Language Models [61.6896704217147]
  Large Vision-Language Models (LVLMs) have become indispensable tools in computer vision and natural language processing. Our investigation reveals a noteworthy bias in the generated content, where the output is primarily influenced by the underlying Large Language Models (LLMs) prior to the input image. To rectify these biases and redirect the model's focus toward vision information, we introduce two simple, training-free strategies.
  arXiv  Detail & Related papers  (2024-03-08T12:35:07Z)
• Mismatch Quest: Visual and Textual Feedback for Image-Text Misalignment [64.49170817854942]
  We present a method to provide detailed explanation of detected misalignments between text-image pairs. We leverage large language models and visual grounding models to automatically construct a training set that holds plausible captions for a given image. We also publish a new human curated test set comprising ground-truth textual and visual misalignment annotations.
  arXiv  Detail & Related papers  (2023-12-05T20:07:34Z)
• PRIME: Prioritizing Interpretability in Failure Mode Extraction [49.93565079216376]
  We study the challenge of providing human-understandable descriptions for failure modes in trained image classification models. We propose a novel approach that prioritizes interpretability in this problem. Our method successfully identifies failure modes and generates high-quality text descriptions associated with them.
  arXiv  Detail & Related papers  (2023-09-29T22:00:12Z)
• Textual Enhanced Contrastive Learning for Solving Math Word Problems [23.196339273292246]
  We propose a Textual Enhanced Contrastive Learning framework, which enforces the models to distinguish semantically similar examples. We adopt a self-supervised manner strategy to enrich examples with subtle textual variance. Experimental results show that our method achieves state-of-the-art on both widely used benchmark datasets and also exquisitely designed challenge datasets in English and Chinese.
  arXiv  Detail & Related papers  (2022-11-29T08:44:09Z)
• Measuring Representational Harms in Image Captioning [5.543867614999908]
  We present a set of techniques for measuring five types of representational harms, as well as the resulting measurements. Our goal was not to audit this image captioning system, but rather to develop normatively grounded measurement techniques. We discuss the assumptions underlying our measurement approach and point out when they do not hold.
  arXiv  Detail & Related papers  (2022-06-14T21:08:01Z)

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.