LAION-SG: An Enhanced Large-Scale Dataset for Training Complex Image-Text Models with Structural Annotations

• URL: http://arxiv.org/abs/2412.08580v2
• Date: Fri, 13 Dec 2024 03:13:47 GMT
• Title: LAION-SG: An Enhanced Large-Scale Dataset for Training Complex Image-Text Models with Structural Annotations
• Authors: Zejian Li, Chenye Meng, Yize Li, Ling Yang, Shengyuan Zhang, Jiarui Ma, Jiayi Li, Guang Yang, Changyuan Yang, Zhiyuan Yang, Jinxiong Chang, Lingyun Sun,
• Abstract summary: Existing text-to-image (T2I) models show decayed performance in compositional image generation involving multiple objects and intricate relationships.<n>We construct LAION-SG, a large-scale dataset with high-quality structural annotations of scene graphs.<n>We also introduce CompSG-Bench, a benchmark that evaluates models on compositional image generation.
• Score: 18.728541981438216
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Recent advances in text-to-image (T2I) generation have shown remarkable success in producing high-quality images from text. However, existing T2I models show decayed performance in compositional image generation involving multiple objects and intricate relationships. We attribute this problem to limitations in existing datasets of image-text pairs, which lack precise inter-object relationship annotations with prompts only. To address this problem, we construct LAION-SG, a large-scale dataset with high-quality structural annotations of scene graphs (SG), which precisely describe attributes and relationships of multiple objects, effectively representing the semantic structure in complex scenes. Based on LAION-SG, we train a new foundation model SDXL-SG to incorporate structural annotation information into the generation process. Extensive experiments show advanced models trained on our LAION-SG boast significant performance improvements in complex scene generation over models on existing datasets. We also introduce CompSG-Bench, a benchmark that evaluates models on compositional image generation, establishing a new standard for this domain. Our annotations with the associated processing code, the foundation model and the benchmark protocol are publicly available at https://github.com/mengcye/LAION-SG.

Related papers

• Hierarchical and Step-Layer-Wise Tuning of Attention Specialty for Multi-Instance Synthesis in Diffusion Transformers [22.269573676129152]
  Text-to-image (T2I) generation models often struggle with multi-instance synthesis (MIS) Traditional MIS control methods for UNet architectures fail to adapt to DiT-based models. We propose a training-free approach for enhancing MIS in DiT-based models.
  arXiv  Detail & Related papers  (2025-04-14T11:59:58Z)
• Interleaved Scene Graph for Interleaved Text-and-Image Generation Assessment [53.45813302866466]
  We present ISG, a comprehensive evaluation framework for interleaved text-and-image generation. ISG evaluates responses on four levels of granularity: holistic, structural, block-level, and image-specific. In conjunction with ISG, we introduce a benchmark, ISG-Bench, encompassing 1,150 samples across 8 categories and 21 subcategories.
  arXiv  Detail & Related papers  (2024-11-26T07:55:57Z)
• SG-Adapter: Enhancing Text-to-Image Generation with Scene Graph Guidance [46.77060502803466]
  We introduce the Scene Graph Adapter(SG-Adapter), leveraging the structured representation of scene graphs to rectify inaccuracies in the original text embeddings. The SG-Adapter's explicit and non-fully connected graph representation greatly improves the fully connected, transformer-based text representations.
  arXiv  Detail & Related papers  (2024-05-24T08:00:46Z)
• AlloyASG: Alloy Predicate Code Representation as a Compact Structurally Balanced Graph [0.6445605125467574]
  We introduce a novel code representation schema, Complex Structurally Balanced Abstract Semantic Graph (CSBASG) CSBASG represents code as a complex-weighted directed graph that lists a semantic element as a node in the graph. We evaluate the efficiency of our CSBASG representation for Alloy models in terms of it's compactness compared to ASTs.
  arXiv  Detail & Related papers  (2024-02-29T22:41:09Z)
• Exploiting Contextual Target Attributes for Target Sentiment Classification [53.30511968323911]
  Existing PTLM-based models for TSC can be categorized into two groups: 1) fine-tuning-based models that adopt PTLM as the context encoder; 2) prompting-based models that transfer the classification task to the text/word generation task. We present a new perspective of leveraging PTLM for TSC: simultaneously leveraging the merits of both language modeling and explicit target-context interactions via contextual target attributes.
  arXiv  Detail & Related papers  (2023-12-21T11:45:28Z)
• LLM Blueprint: Enabling Text-to-Image Generation with Complex and Detailed Prompts [60.54912319612113]
  Diffusion-based generative models have significantly advanced text-to-image generation but encounter challenges when processing lengthy and intricate text prompts. We present a novel approach leveraging Large Language Models (LLMs) to extract critical components from text prompts. Our evaluation on complex prompts featuring multiple objects demonstrates a substantial improvement in recall compared to baseline diffusion models.
  arXiv  Detail & Related papers  (2023-10-16T17:57:37Z)
• SSMG: Spatial-Semantic Map Guided Diffusion Model for Free-form Layout-to-Image Generation [68.42476385214785]
  We propose a novel Spatial-Semantic Map Guided (SSMG) diffusion model that adopts the feature map, derived from the layout, as guidance. SSMG achieves superior generation quality with sufficient spatial and semantic controllability compared to previous works. We also propose the Relation-Sensitive Attention (RSA) and Location-Sensitive Attention (LSA) mechanisms.
  arXiv  Detail & Related papers  (2023-08-20T04:09:12Z)
• T2I-CompBench: A Comprehensive Benchmark for Open-world Compositional Text-to-image Generation [62.71574695256264]
  T2I-CompBench is a comprehensive benchmark for open-world compositional text-to-image generation. We propose several evaluation metrics specifically designed to evaluate compositional text-to-image generation. We introduce a new approach, Generative mOdel fine-tuning with Reward-driven Sample selection (GORS) to boost the compositional text-to-image generation abilities.
  arXiv  Detail & Related papers  (2023-07-12T17:59:42Z)
• LeftRefill: Filling Right Canvas based on Left Reference through Generalized Text-to-Image Diffusion Model [55.20469538848806]
  LeftRefill is an innovative approach to harness large Text-to-Image (T2I) diffusion models for reference-guided image synthesis. This paper introduces LeftRefill, an innovative approach to efficiently harness large Text-to-Image (T2I) diffusion models for reference-guided image synthesis.
  arXiv  Detail & Related papers  (2023-05-19T10:29:42Z)
• Image Inpainting via Conditional Texture and Structure Dual Generation [26.97159780261334]
  We propose a novel two-stream network for image inpainting, which models the structure-constrained texture synthesis and texture-guided structure reconstruction. To enhance the global consistency, a Bi-directional Gated Feature Fusion (Bi-GFF) module is designed to exchange and combine the structure and texture information. Experiments on the CelebA, Paris StreetView and Places2 datasets demonstrate the superiority of the proposed method.
  arXiv  Detail & Related papers  (2021-08-22T15:44:37Z)
• Aggregated Contextual Transformations for High-Resolution Image Inpainting [57.241749273816374]
  We propose an enhanced GAN-based model, named Aggregated COntextual-Transformation GAN (AOT-GAN) for high-resolution image inpainting. To enhance context reasoning, we construct the generator of AOT-GAN by stacking multiple layers of a proposed AOT block. For improving texture synthesis, we enhance the discriminator of AOT-GAN by training it with a tailored mask-prediction task.
  arXiv  Detail & Related papers  (2021-04-03T15:50:17Z)
• Object-Centric Image Generation from Layouts [93.10217725729468]
  We develop a layout-to-image-generation method to generate complex scenes with multiple objects. Our method learns representations of the spatial relationships between objects in the scene, which lead to our model's improved layout-fidelity. We introduce SceneFID, an object-centric adaptation of the popular Fr'echet Inception Distance metric, that is better suited for multi-object images.
  arXiv  Detail & Related papers  (2020-03-16T21:40:09Z)

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.