T2ISafety: Benchmark for Assessing Fairness, Toxicity, and Privacy in Image Generation

• URL: http://arxiv.org/abs/2501.12612v2
• Date: Thu, 20 Feb 2025 11:18:20 GMT
• Title: T2ISafety: Benchmark for Assessing Fairness, Toxicity, and Privacy in Image Generation
• Authors: Lijun Li, Zhelun Shi, Xuhao Hu, Bowen Dong, Yiran Qin, Xihui Liu, Lu Sheng, Jing Shao,
• Abstract summary: T2ISafety is a safety benchmark that evaluates T2I models across three key domains: toxicity, fairness, and bias. We build a large-scale T2I dataset with 68K manually annotated images and train an evaluator capable of detecting critical risks. We evaluate 12 prominent diffusion models on T2ISafety and reveal several concerns including persistent issues with racial fairness, a tendency to generate toxic content, and significant variation in privacy protection across the models.
• Score: 39.45602029655288
• License:
• Abstract: Text-to-image (T2I) models have rapidly advanced, enabling the generation of high-quality images from text prompts across various domains. However, these models present notable safety concerns, including the risk of generating harmful, biased, or private content. Current research on assessing T2I safety remains in its early stages. While some efforts have been made to evaluate models on specific safety dimensions, many critical risks remain unexplored. To address this gap, we introduce T2ISafety, a safety benchmark that evaluates T2I models across three key domains: toxicity, fairness, and bias. We build a detailed hierarchy of 12 tasks and 44 categories based on these three domains, and meticulously collect 70K corresponding prompts. Based on this taxonomy and prompt set, we build a large-scale T2I dataset with 68K manually annotated images and train an evaluator capable of detecting critical risks that previous work has failed to identify, including risks that even ultra-large proprietary models like GPTs cannot correctly detect. We evaluate 12 prominent diffusion models on T2ISafety and reveal several concerns including persistent issues with racial fairness, a tendency to generate toxic content, and significant variation in privacy protection across the models, even with defense methods like concept erasing. Data and evaluator are released under https://github.com/adwardlee/t2i_safety.

Related papers

• Can't See the Forest for the Trees: Benchmarking Multimodal Safety Awareness for Multimodal LLMs [56.440345471966666]
  Multimodal Large Language Models (MLLMs) have expanded the capabilities of traditional language models by enabling interaction through both text and images. This paper introduces MMSafeAware, the first comprehensive multimodal safety awareness benchmark designed to evaluate MLLMs across 29 safety scenarios. MMSafeAware includes both unsafe and over-safety subsets to assess models abilities to correctly identify unsafe content and avoid over-sensitivity that can hinder helpfulness.
  arXiv  Detail & Related papers  (2025-02-16T16:12:40Z)
• Rethinking Bottlenecks in Safety Fine-Tuning of Vision Language Models [25.606641582511106]
  We propose a novel dataset that integrates multi-image inputs with safety Chain-of-Thought (CoT) labels as fine-grained reasoning logic to improve model performance. Our experiments demonstrate that fine-tuning InternVL2.5-8B with MIS significantly outperforms both powerful open-source models and API-based models in challenging multi-image tasks.
  arXiv  Detail & Related papers  (2025-01-30T17:59:45Z)
• SafetyDPO: Scalable Safety Alignment for Text-to-Image Generation [68.07258248467309]
  Text-to-image (T2I) models have become widespread, but their limited safety guardrails expose end users to harmful content and potentially allow for model misuse. Current safety measures are typically limited to text-based filtering or concept removal strategies, able to remove just a few concepts from the model's generative capabilities. We introduce SafetyDPO, a method for safety alignment of T2I models through Direct Preference Optimization (DPO) We train safety experts, in the form of low-rank adaptation (LoRA) matrices, able to guide the generation process away from specific safety-related
  arXiv  Detail & Related papers  (2024-12-13T18:59:52Z)
• Multimodal Pragmatic Jailbreak on Text-to-image Models [43.67831238116829]
  This work introduces a novel type of jailbreak, which triggers T2I models to generate the image with visual text. We benchmark nine representative T2I models, including two close-source commercial models. All tested models suffer from such type of jailbreak, with rates of unsafe generation ranging from 8% to 74%.
  arXiv  Detail & Related papers  (2024-09-27T21:23:46Z)
• Direct Unlearning Optimization for Robust and Safe Text-to-Image Models [29.866192834825572]
  Unlearning techniques have been developed to remove the model's ability to generate potentially harmful content. These methods are easily bypassed by adversarial attacks, making them unreliable for ensuring the safety of generated images. We propose Direct Unlearning Optimization (DUO), a novel framework for removing Not Safe For Work (NSFW) content from T2I models.
  arXiv  Detail & Related papers  (2024-07-17T08:19:11Z)
• Six-CD: Benchmarking Concept Removals for Benign Text-to-image Diffusion Models [58.74606272936636]
  Text-to-image (T2I) diffusion models have shown exceptional capabilities in generating images that closely correspond to textual prompts. The models could be exploited for malicious purposes, such as generating images with violence or nudity, or creating unauthorized portraits of public figures in inappropriate contexts. concept removal methods have been proposed to modify diffusion models to prevent the generation of malicious and unwanted concepts.
  arXiv  Detail & Related papers  (2024-06-21T03:58:44Z)
• PhyBench: A Physical Commonsense Benchmark for Evaluating Text-to-Image Models [50.33699462106502]
  Text-to-image (T2I) models frequently fail to produce images consistent with physical commonsense. Current T2I evaluation benchmarks focus on metrics such as accuracy, bias, and safety, neglecting the evaluation of models' internal knowledge. We introduce PhyBench, a comprehensive T2I evaluation dataset comprising 700 prompts across 4 primary categories: mechanics, optics, thermodynamics, and material properties.
  arXiv  Detail & Related papers  (2024-06-17T17:49:01Z)
• Position: Towards Implicit Prompt For Text-To-Image Models [57.00716011456852]
  This paper highlights the current state of text-to-image (T2I) models toward implicit prompts. We present a benchmark named ImplicitBench and conduct an investigation on the performance and impacts of implicit prompts. Experiment results show that T2I models are able to accurately create various target symbols indicated by implicit prompts.
  arXiv  Detail & Related papers  (2024-03-04T15:21:51Z)
• Adversarial Nibbler: An Open Red-Teaming Method for Identifying Diverse Harms in Text-to-Image Generation [19.06501699814924]
  We build the Adversarial Nibbler Challenge, a red-teaming methodology for crowdsourcing implicitly adversarial prompts. The challenge is run in consecutive rounds to enable a sustained discovery and analysis of safety pitfalls in T2I models. We find that 14% of images that humans consider harmful are mislabeled as safe'' by machines.
  arXiv  Detail & Related papers  (2024-02-14T22:21:12Z)
• Adversarial Nibbler: A Data-Centric Challenge for Improving the Safety of Text-to-Image Models [6.475537049815622]
  Adversarial Nibbler is a data-centric challenge, part of the DataPerf challenge suite, organized and supported by Kaggle and MLCommons.
  arXiv  Detail & Related papers  (2023-05-22T15:02:40Z)

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.