How to Train your Text-to-Image Model: Evaluating Design Choices for Synthetic Training Captions

• URL: http://arxiv.org/abs/2506.16679v1
• Date: Fri, 20 Jun 2025 01:52:17 GMT
• Title: How to Train your Text-to-Image Model: Evaluating Design Choices for Synthetic Training Captions
• Authors: Manuel Brack, Sudeep Katakol, Felix Friedrich, Patrick Schramowski, Hareesh Ravi, Kristian Kersting, Ajinkya Kale,
• Abstract summary: We investigate how different synthetic captioning strategies impact the downstream performance of text-to-image models.<n>Our experiments demonstrate that dense, high-quality captions enhance text alignment but may introduce trade-offs in output aesthetics and diversity.<n>Our findings underscore the importance of caption design in achieving optimal model performance.
• Score: 29.52344052330828
• License: http://creativecommons.org/licenses/by-nc-sa/4.0/
• Abstract: Training data is at the core of any successful text-to-image models. The quality and descriptiveness of image text are crucial to a model's performance. Given the noisiness and inconsistency in web-scraped datasets, recent works shifted towards synthetic training captions. While this setup is generally believed to produce more capable models, current literature does not provide any insights into its design choices. This study closes this gap by systematically investigating how different synthetic captioning strategies impact the downstream performance of text-to-image models. Our experiments demonstrate that dense, high-quality captions enhance text alignment but may introduce trade-offs in output aesthetics and diversity. Conversely, captions of randomized lengths yield balanced improvements across aesthetics and alignment without compromising sample diversity. We also demonstrate that varying caption distributions introduce significant shifts in the output bias of a trained model. Our findings underscore the importance of caption design in achieving optimal model performance and provide practical insights for more effective training data strategies in text-to-image generation.

Related papers

• EDITOR: Effective and Interpretable Prompt Inversion for Text-to-Image Diffusion Models [31.31018600797305]
  We propose a prompt inversion technique called sys for text-to-image diffusion models.<n>Our method outperforms existing methods in terms of image similarity, textual alignment, prompt interpretability and generalizability.
  arXiv  Detail & Related papers  (2025-06-03T16:44:15Z)
• Precision or Recall? An Analysis of Image Captions for Training Text-to-Image Generation Model [32.14771853421448]
  We analyze the critical role of caption precision and recall in text-to-image model training. We utilize Large Vision Language Models to generate synthetic captions for training.
  arXiv  Detail & Related papers  (2024-11-07T19:00:37Z)
• Revisit Large-Scale Image-Caption Data in Pre-training Multimodal Foundation Models [63.01630478059315]
  Recent advancements in multimodal models highlight the value of rewritten captions for improving performance. It is not clear whether synthetic captions and their interaction with original web-crawled AltTexts in pre-training is still not well understood. We propose a novel, controllable, and scalable captioning pipeline designed to generate diverse caption formats tailored to various multimodal models.
  arXiv  Detail & Related papers  (2024-10-03T17:54:52Z)
• Removing Distributional Discrepancies in Captions Improves Image-Text Alignment [76.31530836622694]
  We introduce a model designed to improve the prediction of image-text alignment. Our approach focuses on generating high-quality training datasets for the alignment task. We also demonstrate the applicability of our model by ranking the images generated by text-to-image models based on text alignment.
  arXiv  Detail & Related papers  (2024-10-01T17:50:17Z)
• ARTIST: Improving the Generation of Text-rich Images with Disentangled Diffusion Models and Large Language Models [52.23899502520261]
  We introduce a novel framework named, ARTIST, which incorporates a dedicated textual diffusion model to focus on the learning of text structures specifically.<n>We finetune a visual diffusion model, enabling it to assimilate textual structure information from the pretrained textual model.<n>This disentangled architecture design and training strategy significantly enhance the text rendering ability of the diffusion models for text-rich image generation.
  arXiv  Detail & Related papers  (2024-06-17T19:31:24Z)
• Improving Text Generation on Images with Synthetic Captions [2.1175632266708733]
  latent diffusion models such as SDXL and SD 1.5 have shown significant capability in generating realistic images. We propose a low-cost approach by leveraging SDXL without any time-consuming training on large-scale datasets. Our results demonstrate how our small scale fine-tuning approach can improve the accuracy of text generation in different scenarios.
  arXiv  Detail & Related papers  (2024-06-01T17:27:34Z)
• Training-Free Consistent Text-to-Image Generation [80.4814768762066]
  Text-to-image models can portray the same subject across diverse prompts. Existing approaches fine-tune the model to teach it new words that describe specific user-provided subjects. We present ConsiStory, a training-free approach that enables consistent subject generation by sharing the internal activations of the pretrained model.
  arXiv  Detail & Related papers  (2024-02-05T18:42:34Z)
• Improving Cross-modal Alignment with Synthetic Pairs for Text-only Image Captioning [13.357749288588039]
  Previous works leverage the CLIP's cross-modal association ability for image captioning, relying solely on textual information under unsupervised settings. This paper proposes a novel method to address these issues by incorporating synthetic image-text pairs. A pre-trained text-to-image model is deployed to obtain images that correspond to textual data, and the pseudo features of generated images are optimized toward the real ones in the CLIP embedding space.
  arXiv  Detail & Related papers  (2023-12-14T12:39:29Z)
• Image Captions are Natural Prompts for Text-to-Image Models [70.30915140413383]
  We analyze the relationship between the training effect of synthetic data and the synthetic data distribution induced by prompts. We propose a simple yet effective method that prompts text-to-image generative models to synthesize more informative and diverse training data. Our method significantly improves the performance of models trained on synthetic training data.
  arXiv  Detail & Related papers  (2023-07-17T14:38:11Z)
• WordStylist: Styled Verbatim Handwritten Text Generation with Latent Diffusion Models [8.334487584550185]
  We present a latent diffusion-based method for styled text-to-text-content-image generation on word-level. Our proposed method is able to generate realistic word image samples from different writer styles. We show that the proposed model produces samples that are aesthetically pleasing, help boosting text recognition performance, and get similar writer retrieval score as real data.
  arXiv  Detail & Related papers  (2023-03-29T10:19:26Z)
• Generating More Pertinent Captions by Leveraging Semantics and Style on Multi-Source Datasets [56.018551958004814]
  This paper addresses the task of generating fluent descriptions by training on a non-uniform combination of data sources. Large-scale datasets with noisy image-text pairs provide a sub-optimal source of supervision. We propose to leverage and separate semantics and descriptive style through the incorporation of a style token and keywords extracted through a retrieval component.
  arXiv  Detail & Related papers  (2021-11-24T19:00:05Z)
• 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.