Evaluating Image Hallucination in Text-to-Image Generation with Question-Answering

• URL: http://arxiv.org/abs/2409.12784v4
• Date: Tue, 15 Oct 2024 15:01:44 GMT
• Title: Evaluating Image Hallucination in Text-to-Image Generation with Question-Answering
• Authors: Youngsun Lim, Hojun Choi, Hyunjung Shim,
• Abstract summary: We introduce I-HallA (Image Hallucination evaluation with Question Answering), a novel evaluation metric. I-HallA measures the factuality of generated images through visual question answering (VQA) We evaluate five text-to-image models using I-HallA and reveal that these state-of-the-art models often fail to accurately convey factual information.
• Score: 13.490305443938817
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Despite the impressive success of text-to-image (TTI) generation models, existing studies overlook the issue of whether these models accurately convey factual information. In this paper, we focus on the problem of image hallucination, where images created by generation models fail to faithfully depict factual content. To address this, we introduce I-HallA (Image Hallucination evaluation with Question Answering), a novel automated evaluation metric that measures the factuality of generated images through visual question answering (VQA). We also introduce I-HallA v1.0, a curated benchmark dataset for this purpose. As part of this process, we develop a pipeline that generates high-quality question-answer pairs using multiple GPT-4 Omni-based agents, with human judgments to ensure accuracy. Our evaluation protocols measure image hallucination by testing if images from existing text-to-image models can correctly respond to these questions. The I-HallA v1.0 dataset comprises 1.2K diverse image-text pairs across nine categories with 1,000 rigorously curated questions covering various compositional challenges. We evaluate five text-to-image models using I-HallA and reveal that these state-of-the-art models often fail to accurately convey factual information. Moreover, we validate the reliability of our metric by demonstrating a strong Spearman correlation (rho=0.95) with human judgments. We believe our benchmark dataset and metric can serve as a foundation for developing factually accurate text-to-image generation models.

Related papers

• D-Judge: How Far Are We? Evaluating the Discrepancies Between AI-synthesized Images and Natural Images through Multimodal Guidance [19.760989919485894]
  We introduce an AI-Natural Image Discrepancy accessing benchmark (textitD-Judge) We construct textitD-ANI, a dataset with 5,000 natural images and over 440,000 AIGIs generated by nine models using Text-to-Image (T2I), Image-to-Image (I2I), and Text and Image-to-Image (TI2I) prompts. Our framework evaluates the discrepancy across five dimensions: naive image quality, semantic alignment, aesthetic appeal, downstream applicability, and human validation.
  arXiv  Detail & Related papers  (2024-12-23T15:08:08Z)
• Evaluating Hallucination in Text-to-Image Diffusion Models with Scene-Graph based Question-Answering Agent [9.748808189341526]
  An effective Text-to-Image (T2I) evaluation metric should accomplish the following: detect instances where the generated images do not align with the textual prompts. We propose a method based on large language models (LLMs) for conducting question-answering with an extracted scene-graph and created a dataset with human-rated scores for generated images.
  arXiv  Detail & Related papers  (2024-12-07T18:44:38Z)
• Image2Text2Image: A Novel Framework for Label-Free Evaluation of Image-to-Text Generation with Text-to-Image Diffusion Models [16.00576040281808]
  We propose a novel framework called Image2Text2Image to evaluate image captioning models. A high similarity score suggests that the model has produced a faithful textual description, while a low score highlights discrepancies. Our framework does not rely on human-annotated captions reference, making it a valuable tool for assessing image captioning models.
  arXiv  Detail & Related papers  (2024-11-08T17:07:01Z)
• How Many Van Goghs Does It Take to Van Gogh? Finding the Imitation Threshold [50.33428591760124]
  We study the relationship between a concept's frequency in the training dataset and the ability of a model to imitate it. We propose an efficient approach that estimates the imitation threshold without incurring the colossal cost of training multiple models from scratch.
  arXiv  Detail & Related papers  (2024-10-19T06:28:14Z)
• KITTEN: A Knowledge-Intensive Evaluation of Image Generation on Visual Entities [93.74881034001312]
  We conduct a systematic study on the fidelity of entities in text-to-image generation models. We focus on their ability to generate a wide range of real-world visual entities, such as landmark buildings, aircraft, plants, and animals. Our findings reveal that even the most advanced text-to-image models often fail to generate entities with accurate visual details.
  arXiv  Detail & Related papers  (2024-10-15T17:50:37Z)
• A Novel Evaluation Framework for Image2Text Generation [15.10524860121122]
  We propose an evaluation framework rooted in a modern large language model (LLM) capable of image generation. A high similarity score suggests that the image captioning model has accurately generated textual descriptions. A low similarity score indicates discrepancies, revealing potential shortcomings in the model's performance.
  arXiv  Detail & Related papers  (2024-08-03T09:27:57Z)
• Holistic Evaluation of Text-To-Image Models [153.47415461488097]
  We introduce a new benchmark, Holistic Evaluation of Text-to-Image Models (HEIM) We identify 12 aspects, including text-image alignment, image quality, aesthetics, originality, reasoning, knowledge, bias, toxicity, fairness, robustness, multilinguality, and efficiency. Our results reveal that no single model excels in all aspects, with different models demonstrating different strengths.
  arXiv  Detail & Related papers  (2023-11-07T19:00:56Z)
• Davidsonian Scene Graph: Improving Reliability in Fine-grained Evaluation for Text-to-Image Generation [64.64849950642619]
  We develop an evaluation framework inspired by formal semantics for evaluating text-to-image models. We show that Davidsonian Scene Graph (DSG) produces atomic and unique questions organized in dependency graphs. We also present DSG-1k, an open-sourced evaluation benchmark that includes 1,060 prompts.
  arXiv  Detail & Related papers  (2023-10-27T16:20:10Z)
• On quantifying and improving realism of images generated with diffusion [50.37578424163951]
  We propose a metric, called Image Realism Score (IRS), computed from five statistical measures of a given image. IRS is easily usable as a measure to classify a given image as real or fake. We experimentally establish the model- and data-agnostic nature of the proposed IRS by successfully detecting fake images generated by Stable Diffusion Model (SDM), Dalle2, Midjourney and BigGAN. Our efforts have also led to Gen-100 dataset, which provides 1,000 samples for 100 classes generated by four high-quality models.
  arXiv  Detail & Related papers  (2023-09-26T08:32:55Z)
• StableRep: Synthetic Images from Text-to-Image Models Make Strong Visual Representation Learners [58.941838860425754]
  We show that training self-supervised methods on synthetic images can match or beat the real image counterpart. We develop a multi-positive contrastive learning method, which we call StableRep. With solely synthetic images, the representations learned by StableRep surpass the performance of representations learned by SimCLR and CLIP.
  arXiv  Detail & Related papers  (2023-06-01T17:59:51Z)
• TIFA: Accurate and Interpretable Text-to-Image Faithfulness Evaluation with Question Answering [86.38098280689027]
  We introduce an automatic evaluation metric that measures the faithfulness of a generated image to its text input via visual question answering (VQA) We present a comprehensive evaluation of existing text-to-image models using a benchmark consisting of 4K diverse text inputs and 25K questions across 12 categories (object, counting, etc.)
  arXiv  Detail & Related papers  (2023-03-21T14:41:02Z)
• How good are deep models in understanding the generated images? [47.64219291655723]
  Two sets of generated images are collected for object recognition and visual question answering tasks. On object recognition, the best model, out of 10 state-of-the-art object recognition models, achieves about 60% and 80% top-1 and top-5 accuracy. On VQA, the OFA model scores 77.3% on answering 241 binary questions across 50 images.
  arXiv  Detail & Related papers  (2022-08-23T06:44:43Z)
• Improving Generation and Evaluation of Visual Stories via Semantic Consistency [72.00815192668193]
  Given a series of natural language captions, an agent must generate a sequence of images that correspond to the captions. Prior work has introduced recurrent generative models which outperform synthesis text-to-image models on this task. We present a number of improvements to prior modeling approaches, including the addition of a dual learning framework.
  arXiv  Detail & Related papers  (2021-05-20T20:42:42Z)

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.