mlsanityData to 14 Oct 2026

XVerse: Consistent Multi-Subject Control of Identity and Semantic Attributes via DiT Modulation

arXiv 2506.21416 ↗First posted 26 Jun 2025cs.CVcited by 56

On the map: Diffusion models → Image generation, text-to-image and editing

Abstract (arXiv)

Achieving fine-grained control over subject identity and semantic attributes (pose, style, lighting) in text-to-image generation, particularly for multiple subjects, often undermines the editability and coherence of Diffusion Transformers (DiTs). Many approaches introduce artifacts or suffer from attribute entanglement. To overcome these challenges, we propose a novel multi-subject controlled generation model XVerse. By transforming reference images into offsets for token-specific text-stream modulation, XVerse allows for precise and independent control for specific subject without disrupting image latents or features. Consequently, XVerse offers high-fidelity, editable multi-subject image synthesis with robust control over individual subject characteristics and semantic attributes. This advancement significantly improves personalized and complex scene generation capabilities.

Key facts. Cited by 56 papers in the map (data through 14 Oct 2026). Placed in Image generation, text-to-image and editing, one of 14 directions in Diffusion models. Builds on 32 of its 40 references that are in the map. First posted on arXiv 26 Jun 2025.

Builds on

32 of 40 references.

Closest ancestors · 6

Same direction, newest first

  1. 2025TokenVerse: Versatile Multi-concept Personalization in Token Modulation Space21 Jan
  2. 2024OmniGen: Unified Image Generation17 Sep
  3. 2024ELLA: Equip Diffusion Models with LLM for Enhanced Semantic Alignment8 Mar
  4. 2023PhotoVerse: Tuning-Free Image Customization with Text-to-Image Diffusion Models11 Sep
  5. 2023IP-Adapter: Text Compatible Image Prompt Adapter for Text-to-Image Diffusion Models13 Aug
  6. 2022DreamBooth: Fine Tuning Text-to-Image Diffusion Models for Subject-Driven Generation25 Aug

Other references · 26

  1. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,744
  2. 2020Denoising Diffusion Probabilistic Modelscited by 21,076
  3. 2023DINOv2: Learning Robust Visual Features without Supervisioncited by 7,974
  4. 2015Deep Unsupervised Learning using Nonequilibrium Thermodynamicscited by 7,058
  5. 2022Scalable Diffusion Models with Transformerscited by 6,054
  6. 2018A Style-Based Generator Architecture for Generative Adversarial Networkscited by 6,013
  7. 2022Flamingo: a Visual Language Model for Few-Shot Learningcited by 4,386
  8. 2024Scaling Rectified Flow Transformers for High-Resolution Image Synthesiscited by 4,181
  9. 2024SAM 2: Segment Anything in Images and Videoscited by 2,863
  10. 2017FiLM: Visual Reasoning with a General Conditioning Layercited by 2,333
  11. 2022An Image is Worth One Word: Personalizing Text-to-Image Generation using Textual Inversioncited by 2,245
  12. 2021LAION-400M: Open Dataset of CLIP-Filtered 400 Million Image-Text Pairscited by 1,555
14 more
  1. 2020Encoding in Style: a StyleGAN Encoder for Image-to-Image Translationcited by 727
  2. 2020GANSpace: Discovering Interpretable GAN Controlscited by 614
  3. 2021Designing an Encoder for StyleGAN Image Manipulationcited by 544
  4. 2021Pivotal Tuning for Latent-based Editing of Real Imagescited by 427
  5. 2020In-Domain GAN Inversion for Real Image Editingcited by 420
  6. 2023Florence-2: Advancing a Unified Representation for a Variety of Vision Taskscited by 410
  7. 2024OminiControl: Minimal and Universal Control for Diffusion Transformercited by 340
  8. 2025Less-to-More Generalization: Unlocking More Controllability by In-Context Generationcited by 154
  9. 2024PuLID: Pure and Lightning ID Customization via Contrastive Alignmentcited by 144
  10. 2024MS-Diffusion: Multi-subject Zero-shot Image Personalization with Layout Guidancecited by 144
  11. 2024DreamBench++: A Human-Aligned Benchmark for Personalized Image Generationcited by 131
  12. 2024UniReal: Universal Image Generation and Editing via Learning Real-world Dynamicscited by 102
  13. 2025DreamO: A Unified Framework for Image Customizationcited by 83
  14. 2024Resolving Multi-Condition Confusion for Finetuning-Free Personalized Image Generationcited by 43

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 26; 2026: 30.

56 papers.

Sort:

By area:

  1. 2025DreamOmni2: Multimodal Instruction-based Editing and Generationcited by 38
  2. 2025UMO: Scaling Multi-Identity Consistency for Image Customization via Matching Rewardcited by 19
  3. 2025ContextGen: Contextual Layout Anchoring for Identity-Consistent Multi-Instance Generationcited by 18
  4. 2025WithAnyone: Towards Controllable and ID Consistent Image Generationcited by 17
  5. 2025MOSAIC: Multi-Subject Personalized Generation via Correspondence-Aware Alignment and Disentanglementcited by 14
  6. 2025MICo-150K: A Comprehensive Dataset Advancing Multi-Image Compositioncited by 10
  7. 2025Scone: Bridging Composition and Distinction in Subject-Driven Image Generation via Unified Understanding-Generation Modelingcited by 10
  8. 2025PSR: Scaling Multi-Subject Personalized Image Generation with Pairwise Subject-Consistency Rewardscited by 9
  9. 2025EditIDv2: Editable ID Customization with Data-Lubricated ID Feature Integration for Text-to-Image Generationcited by 8
  10. 2025LAMIC: Layout-Aware Multi-Image Composition via Scalability of Multimodal Diffusion Transformercited by 7
  11. 2026OmniTransfer: All-in-one Framework for Spatio-temporal Video Transfercited by 6
  12. 2025SIGMA-GEN: Structure and Identity Guided Multi-subject Assembly for Image Generationcited by 5
38 more
  1. 2025OmniRefiner: Reinforcement-Guided Local Diffusion Refinementcited by 5
  2. 2025The Consistency Critic: Correcting Inconsistencies in Generated Images via Reference-Guided Attentive Alignmentcited by 5
  3. 2026MICON-Bench: Benchmarking and Enhancing Multi-Image Context Image Generation in Unified Multimodal Modelscited by 5
  4. 2025MultiCrafter: High-Fidelity Multi-Subject Generation via Disentangled Attention and Identity-Aware Preference Alignmentcited by 4
  5. 2025OpenSubject: Leveraging Video-Derived Identity and Diversity Priors for Subject-driven Image Generation and Manipulationcited by 4
  6. 2026Garments2Look: A Multi-Reference Dataset for High-Fidelity Outfit-Level Virtual Try-On with Clothing and Accessoriescited by 4
  7. 2026MultiBind: A Benchmark for Attribute Misbinding in Multi-Subject Generationcited by 4
  8. 2025FocusDPO: Dynamic Preference Optimization for Multi-Subject Personalized Image Generation via Adaptive Focuscited by 3
  9. 2025DVI: Disentangling Semantic and Visual Identity for Training-Free Personalized Generationcited by 3
  10. 2026FlexID: Training-Free Flexible Identity Injection via Intent-Aware Modulation for Text-to-Image Generationcited by 3
  11. 2026Premier: Personalized Preference Modulation with Learnable User Embedding in Text-to-Image Generationcited by 3
  12. 2026Scaling Multi-Reference Image Generation with Dynamic Reward Optimizationcited by 3
  13. 2025Ar2Can: An Architect and an Artist Leveraging a Canvas for Multi-Human Generationcited by 2
  14. 2025Composing Concepts from Images and Videos via Concept-prompt Bindingcited by 2
  15. 2025DreamOmni3: Scribble-based Editing and Generationcited by 2
  16. 2026SIGMA: Selective-Interleaved Generation with Multi-Attribute Tokenscited by 2
  17. 2026Composing Driving Worlds through Disentangled Control for Adversarial Scenario Generationcited by 2
  18. 2026When Identities Collapse: A Stress-Test Benchmark for Multi-Subject Personalizationcited by 2
  19. 2026ConceptWeaver: Weaving Disentangled Concepts with Flowcited by 2
  20. 2026UniCustom: Unified Visual Conditioning for Multi-Reference Image Generationcited by 2
  21. 2026MIBE: Multi-subject Interaction Benchmark and Evaluator for Personalized Image Generationcited by 2
  22. 2025DiffInk: Glyph- and Style-Aware Latent Diffusion Transformer for Text to Online Handwriting Generationcited by 1
  23. 2025Adversarial Concept Distillation for One-Step Diffusion Personalizationcited by 1
  24. 2025ContextDrag: Precise Drag-Based Image Editing via Context-Preserving Token Injection and Position-Aligned Attentioncited by 1
  25. 2025Dynamic Image Prompt Adapter for Scalable Zero-shot Personalized Text-to-Image Generationcited by 1
  26. 2026Hierarchical Concept-to-Appearance Guidance for Multi-Subject Image Generationcited by 1
  27. 2026DreamShot: Personalized Storyboard Synthesis with Video Diffusion Priorcited by 1
  28. 2026CogCanvas: A Benchmark for Evaluating Multi-Subject Reference-Based Image Generationcited by 1
  29. 2026StructGen: Disambiguating Multi-Reference Image Generation via Structured Context Modelingcited by 1
  30. 2026MIEScore: Human-Aligned Evaluation for Multi-Source Image Editingcited by 1
  31. 2025FreeFuse: Multi-Subject LoRA Fusion via Adaptive Token-Level Routing at Test Timecited by 0
  32. 2025InstructMoLE: Instruction-Guided Mixture of Low-rank Experts for Multi-Conditional Image Generationcited by 0
  33. 2026MoGen: A Unified Collaborative Framework for Controllable Multi-Object Image Generationcited by 0
  34. 2026AnyPhoto: Multi-Person Identity Preserving Image Generation with ID Adaptive Modulation on Location Canvascited by 0
  35. 2026Training-Free Image Editing with Visual Context Integration and Concept Alignmentcited by 0
  36. 2026Squeezing Capacity from Multimodal Large Language Models for Subject-driven Generationcited by 0
  37. 2026UniVerse: A Unified Modulation Framework for Segmentation-Free,Disentangled Multi-Concept Personalizationcited by 0
  38. 2026CopyCat: Improving Fine-Grained Subject Consistency in Subject-to-Image Models within Secondscited by 0

50 most cited of 56 All 56 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2026MS-CustomNet: Controllable Multi-Subject Customization with Hierarchical Relational Semantics
  2. 2023Subject-Diffusion:Open Domain Personalized Text-to-Image Generation without Test-time Fine-tuning
  3. 2025Personalize Anything for Free with Diffusion Transformer
  4. 2025ComposeMe: Attribute-Specific Image Prompts for Controllable Human Image Generation
  5. 2026Disentangling to Re-couple: Resolving the Similarity-Controllability Paradox in Subject-Driven Text-to-Image Generation
  6. 2025Enhancing Image Generation Fidelity via Progressive Prompts
  7. 2024ComFusion: Personalized Subject Generation in Multiple Specific Scenes From Single Image
  8. 2024Towards a Simultaneous and Granular Identity-Expression Control in Personalized Face Generation