Squeezing Capacity from Multimodal Large Language Models for Subject-driven Generation
cs.CV, cs.AI, cs.GR, cs.LG, cs.MM
Submitted: 2026-05-25
Updated: 2026-11-07
Code: https://github.com/black-forest-labs/flux
Project page: https://zsh2000.github.io/squeeze-mllm-subject-gen
Terminology
Sources
- Qwen3-VL Technical Report
- Qwen2.5-VL Technical Report
- Canvas-to-Image: Compositional Image Generation with Multimodal Controls
- DreamArtist++: Controllable One-Shot Text-to-Image Generation via Positive-Negative Adapter
- TBStar-Edit: From Image Editing Pattern Shifting to Consistency Enhancement
- iMontage: Unified, Versatile, Highly Dynamic Many-to-many Image Generation
- Planting a SEED of Vision in Large Language Model
- Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities
- Gemma 3 Technical Report
- MUSAR: Exploring Multi-Subject Customization from Single-Subject Dataset via Attention Routing
- Re-Align: Structured Reasoning-guided Alignment for In-Context Image Generation and Editing
- In-Context LoRA for Diffusion Transformers
- From Competition to Synergy: Unlocking Reinforcement Learning for Subject-Driven Image Generation
- Taming Encoder for Zero Fine-tuning Image Customization with Text-to-Image Diffusion Models
- EditIDv2: Editable ID Customization with Data-Lubricated ID Feature Integration for Text-to-Image Generation
- DVI: Disentangling Semantic and Visual Identity for Training-Free Personalized Generation
- Inject Where It Matters: Training-Free Spatially-Adaptive Identity Preservation for Text-to-Image Personalization
- UniFusion: Vision-Language Model as Unified Encoder in Image Generation
- Mini-Gemini: Mining the Potential of Multi-modality Vision Language Models
- IC-Custom: Diverse Image Customization via In-Context Learning
Related papers
- TEA: Text Encoder Alignment for Robust Concept Erasure in Text-to-Image Models
- Do Egocentric Video-Language Models Capture Both Hand- and Object-Centric Cues?
- Aura: Consistent Multi-Subject Video Generation via VLM-Grounded Semantic Alignment
- PVRF: All-in-one Adverse Weather Removal via Prior-modulated and Velocity-constrained Rectified Flow
- DLWM: Diverse Latent World Models for Efficient Multimodal Reasoning
- Out of Sight, Still in Mind: Token Compression for Omni-LLMs