A safety direction estimated in a source LLM is transported to a target generator through lightweight alignment on benign data alone, matching native safety performance without any target-side unsafe data.
super hub Mixed citations
Qwen-Image Technical Report
Mixed citation behavior. Most common role is background (46%).
347
Pith papers citing it
Background
46% of classified citations
abstract
We present Qwen-Image, an image generation foundation model in the Qwen series that achieves significant advances in complex text rendering and precise image editing. To address the challenges of complex text rendering, we design a comprehensive data pipeline that includes large-scale data collection, filtering, annotation, synthesis, and balancing. Moreover, we adopt a progressive training strategy that starts with non-text-to-text rendering, evolves from simple to complex textual inputs, and gradually scales up to paragraph-level descriptions. This curriculum learning approach substantially enhances the model's native text rendering capabilities. As a result, Qwen-Image not only performs exceptionally well in alphabetic languages such as English, but also achieves remarkable progress on more challenging logographic languages like Chinese. To enhance image editing consistency, we introduce an improved multi-task training paradigm that incorporates not only traditional text-to-image (T2I) and text-image-to-image (TI2I) tasks but also image-to-image (I2I) reconstruction, effectively aligning the latent representations between Qwen2.5-VL and MMDiT. Furthermore, we separately feed the original image into Qwen2.5-VL and the VAE encoder to obtain semantic and reconstructive representations, respectively. This dual-encoding mechanism enables the editing module to strike a balance between preserving semantic consistency and maintaining visual fidelity. Qwen-Image achieves state-of-the-art performance, demonstrating its strong capabilities in both image generation and editing across multiple benchmarks.
hub tools
citation-role summary
background 34
baseline 23
method 10
dataset 2
citation-polarity summary
claims ledger
- abstract We present Qwen-Image, an image generation foundation model in the Qwen series that achieves significant advances in complex text rendering and precise image editing. To address the challenges of complex text rendering, we design a comprehensive data pipeline that includes large-scale data collection, filtering, annotation, synthesis, and balancing. Moreover, we adopt a progressive training strategy that starts with non-text-to-text rendering, evolves from simple to complex textual inputs, and gradually scales up to paragraph-level descriptions. This curriculum learning approach substantially
authors
co-cited works
representative citing papers
The work creates the first dataset and baseline for generating emission textures on 3D objects to reproduce glowing materials from input images.
OP-GRPO is the first off-policy GRPO method for flow-matching models that reuses trajectories via replay buffer and importance sampling corrections, matching on-policy performance with 34.2% of the training steps.
Cornfigurator is the first automated deployment planner for generic any-to-any multimodal models that explores the full range of colocation-to-disaggregation strategies and delivers 1.12x to 6.32x higher goodput than existing systems or expert plans.
OrbitQuant is a data-agnostic PTQ technique for DiTs that uses RPBH rotation in a normalized basis to enable a single codebook across all inputs, achieving SOTA low-bit performance on FLUX.1, CogVideoX and similar models.
Introduces VICIS task and training framework for inferring visual concepts from image sets, with experiments showing better accuracy, diversity, and generalization than standard VLMs on synthetic and ImageNet data.
A new pipeline using canonical LoRAs for view synthesis, deformable 3D Gaussian splatting anchored on D-SMAL, and generative repair to produce animatable 3D dogs from single wild images without 3D supervision.
Introduces VidPair-Halluc benchmark of 1K background-controlled adversarial video pairs and 11K QA pairs generated via PairFlow pipeline to evaluate hallucination in LVMs.
SharpMoE is a plug-and-play post-training method that uses clean latent features and a trajectory routing loss to enable accurate saliency-based routing in diffusion MoE models for improved visual generation.
New 3DLP benchmark with real-world 1K HDR pairs shows state-of-the-art image editing models vary in physical lighting consistency, with best models close to reality but error-prone in low-light regions.
PhyEditBench is a new benchmark for physics-aware image editing with real and synthetic instances plus a training-free PhyWorld baseline that uses test-time scaling to outperform SOTA models.
TryOnCrafter is the first DiT-based framework for camera-controllable video virtual try-on via a renderable 4D try-on proxy distilled from 2D priors into 3DGS avatar animated with SMPL-X.
Introduces 2M synthetic WATER-S dataset and WATERec model achieving 90.40% accuracy on WordArt-Bench, outperforming prior STR methods and VLMs.
Forensic Knowledge Graphs integrate forensic traces, causal dependencies, and scene links via a new authentication network and Iterative Context Refinement to outperform standard detectors and VLMs on detection, localization, and justification.
REALM is the first unified red-teaming benchmark for physical-world VLMs that aligns diverse attack methods via an agentic target-generation pipeline and evaluates them on shared datasets showing text/typographic attacks as most effective.
Sparse Context achieves 2-4x faster inference in reference-conditioned diffusion models by fine-tuning with random token dropping and applying task-aware selection at inference time, without loss of visual quality.
A technique for controllable diversity in text-to-image generation by inducing structured semantic variations at the prompt level via VLM and agentic workflow.
DiT-Reward converts pretrained DiT models into reward predictors that outperform HPSv3 on four benchmarks while providing 1.65x inference speedup.
RS-Gen proposes a plug-and-play agentic framework with a closed-loop reasoning mechanism that augments base image models to achieve SOTA results on WISE Verified and RISEBench.
InterleaveThinker is the first multi-agent pipeline enabling interleaved generation in any image generator through planner-critic agents, SFT on custom datasets, and GRPO RL with accuracy and step-wise rewards.
CineOrchestra unifies control of subjects, events, cameras, and shot transitions in cinematic video generation through entity-centric conditioning primitives and parameter-free coordinated rotary embeddings.
The paper constructs the VIBE benchmark and evaluates six visual in-context learning models on 14 datasets, 12 tasks, and 106 combinations under a unified one-shot protocol, revealing limitations and failure modes.
ImageTime is a benchmark that probes image generation models' visual world modeling by requiring coherent four-state sequences in single images, scored via VLM judge.
SSR-Merge merges LoRAs via subspace construction, inverse correlation decorrelation, and directional steering, shown to match the OLS solution with a streaming implementation that outperforms prior merging methods.
citing papers explorer
-
Uni-Edit: Intelligent Editing Is A General Task For Unified Model Tuning
Uni-Edit introduces a data synthesis pipeline turning VQA data into reasoning-intensive editing instructions, enabling single-task tuning that boosts all three capabilities in models like BAGEL and Janus-Pro.
-
Linear-DPO: Linear Direct Preference Optimization for Diffusion and Flow-Matching Generative Models
Linear-DPO replaces sigmoid utility with linear utility and adds EMA reference to improve preference alignment in diffusion and flow-matching text-to-image models.
-
GeoX: Mastering Geospatial Reasoning Through Self-Play and Verifiable Rewards
GeoX is a self-play RL framework in which a single multimodal policy proposes and solves spatial problems as executable programs over image primitives, using verifiable rewards to improve base VLMs by up to 5.5 points without large curated data.
-
ReAlign: Generalizable Image Forgery Detection via Reasoning-Aligned Representation
ReAlign distills LLM-generated reasoning texts into a lightweight AIGI forgery detector via contrastive image-text alignment to improve generalization on complex forgeries.
-
From Plans to Pixels: Learning to Plan and Orchestrate for Open-Ended Image Editing
A planner-orchestrator system learns long-horizon image editing by maximizing outcome-based rewards from a vision-language judge and refining plans from successful trajectories.
-
Editor's Choice: Evaluating Abstract Intent in Image Editing through Atomic Entity Analysis
Presents Entity-Rubrics and AbstractEdit benchmark to measure image editing models on abstract intent, finding standard models struggle to balance edit intent with image preservation.
-
Edit-Compass & EditReward-Compass: A Unified Benchmark for Image Editing and Reward Modeling
Edit-Compass and EditReward-Compass are new unified benchmarks for fine-grained image editing evaluation and realistic reward modeling in reinforcement learning optimization.
-
Asymmetric Flow Models
AsymFlow uses rank-asymmetric velocity prediction to reach 1.57 FID on ImageNet 256x256 and enables finetuning of latent flow models into superior pixel-space text-to-image generators.
-
Inline Critic Steers Image Editing
Inline Critic uses a learnable token to critique and steer a frozen image-editing model's intermediate layers during generation, delivering state-of-the-art results on GEdit-Bench, RISEBench, and KRIS-Bench.
-
Images in Sentences: Scaling Interleaved Instructions for Unified Visual Generation
INSET embeds images as native tokens in interleaved instructions, outperforming prior methods on multi-image consistency and text alignment as complexity grows.
-
Beyond Text Prompts: Visual-to-Visual Generation as A Unified Paradigm
Proposes V2V-Zero, a training-free framework replacing text conditioning with VLM final-layer hidden states from visual pages, achieving 0.85 on GenEval and 32.7/100 on new Simple-V2V Bench across models including video extension.
-
Generate "Normal", Edit Poisoned: Branding Injection via Hint Embedding in Image Editing
Invisible hints such as logos embedded in images are re-rendered by diffusion models during text-guided editing, enabling phishing and model-poisoning attacks with average success rates of 44.4% and 32.2%.
-
What Concepts Lie Within? Detecting and Suppressing Risky Content in Diffusion Transformers
A method using attention head vectors detects and suppresses risky content generation in Diffusion Transformers at inference time.
-
EditRefiner: A Human-Aligned Agentic Framework for Image Editing Refinement
EditRefiner uses a perception-reasoning-action-evaluation agent loop and the EditFHF-15K human feedback dataset to refine text-guided image edits more accurately than prior methods.
-
Concept-Based Abductive and Contrastive Explanations for Behaviors of Vision Models
Concept-based abductive and contrastive explanations find minimal high-level concepts that causally determine vision model outcomes on individual images or groups sharing a specified behavior.
-
Sparkle: Realizing Lively Instruction-Guided Video Background Replacement via Decoupled Guidance
Sparkle supplies a large-scale dataset and benchmark for instruction-driven video background replacement, enabling models that generate more natural and temporally consistent new scenes than earlier approaches.
-
Secure Seed-Based Multi-bit Watermarking for Diffusion Models from First Principles
A theoretical framework decouples diffusion model generation from watermark decisions, enabling SSB to reach any security-robustness-fidelity regime without model-specific empirical tests.
-
CrossCult-KIBench: A Benchmark for Cross-Cultural Knowledge Insertion in MLLMs
CrossCult-KIBench is a new benchmark for evaluating cross-cultural knowledge insertion in MLLMs, paired with the MCKI baseline method, showing current approaches fail to balance adaptation and preservation.
-
MULTITEXTEDIT: Benchmarking Cross-Lingual Degradation in Text-in-Image Editing
MULTITEXTEDIT benchmark reveals that all tested text-in-image editing models show pronounced degradation on non-English languages, especially Hebrew and Arabic, mainly in text accuracy and script fidelity.
-
SpecEdit: Training-Free Acceleration for Diffusion based Image Editing via Semantic Locking
SpecEdit accelerates diffusion-based image editing up to 10x by using a low-resolution draft to identify edit-relevant tokens via semantic discrepancies for selective high-resolution denoising.
-
VeraRetouch: A Lightweight Fully Differentiable Framework for Multi-Task Reasoning Photo Retouching
VeraRetouch is a 0.5B VLM-based framework with a differentiable Retouch Renderer and a new million-scale AetherRetouch-1M+ dataset that claims state-of-the-art results in reasoning photo retouching while enabling mobile deployment.
-
StyleID: A Perception-Aware Dataset and Metric for Stylization-Agnostic Facial Identity Recognition
StyleID supplies human-perception-aligned benchmarks and fine-tuned encoders that improve facial identity recognition robustness across stylization types and strengths.
-
Exploring Spatial Intelligence from a Generative Perspective
Fine-tuning multimodal models on a new synthetic spatial benchmark improves generative spatial compliance on real and synthetic tasks and transfers to better spatial understanding.
-
ReImagine: Rethinking Controllable High-Quality Human Video Generation via Image-First Synthesis
ReImagine decouples human appearance from temporal consistency via pretrained image backbones, SMPL-X motion guidance, and training-free video diffusion refinement to generate high-quality controllable videos.
-
HP-Edit: A Human-Preference Post-Training Framework for Image Editing
HP-Edit introduces a post-training framework and RealPref-50K dataset that uses a VLM-based HP-Scorer to align diffusion image editing models with human preferences, improving outputs on Qwen-Image-Edit-2509.
-
Long-Text-to-Image Generation via Compositional Prompt Decomposition
PRISM lets pre-trained text-to-image models handle long prompts by breaking them into compositional parts, predicting noise separately, and merging outputs via energy-based conjunction, matching fine-tuned models while generalizing better to prompts over 500 tokens.
-
View-Consistent 3D Scene Editing via Dual-Path Structural Correspondense and Semantic Continuity
A dual-path consistency framework for text-driven 3D scene editing that models cross-view dependencies via structural correspondence and semantic continuity, trained on a newly constructed paired multi-view dataset.
-
UniEditBench: A Unified and Cost-Effective Benchmark for Image and Video Editing via Distilled MLLMs
UniEditBench unifies image and video editing evaluation with a nine-plus-eight operation taxonomy and cost-effective 4B/8B distilled MLLM evaluators that align with human judgments.
-
LeapAlign: Post-Training Flow Matching Models at Any Generation Step by Building Two-Step Trajectories
LeapAlign fine-tunes flow matching models by constructing two consecutive leaps that skip multiple ODE steps with randomized timesteps and consistency weighting, enabling stable updates at any generation step.
-
Beyond Visual Cues: Semantic-Driven Token Filtering and Expert Routing for Anytime Person ReID
STFER uses LVLM-generated identity-consistent semantic text to drive visual token filtering and expert routing for improved any-time person re-identification under clothing changes and modality shifts.
-
LottieGPT: Tokenizing Vector Animation for Autoregressive Generation
LottieGPT tokenizes Lottie animations into compact sequences and fine-tunes Qwen-VL to autoregressively generate coherent vector animations from natural language or visual prompts, outperforming prior SVG models.
-
HiddenObjects: Scalable Diffusion-Distilled Spatial Priors for Object Placement
A diffusion-based pipeline creates a 27M-annotation dataset of object placements that outperforms human annotations and baselines on image editing tasks, then distills it into a fast model.
-
RewardFlow: Generate Images by Optimizing What You Reward
RewardFlow unifies differentiable rewards including a new VQA-based one and uses a prompt-aware adaptive policy with Langevin dynamics to achieve state-of-the-art image editing and compositional generation.
-
FlowGuard: Towards Lightweight In-Generation Safety Detection for Diffusion Models via Linear Latent Decoding
FlowGuard detects unsafe content during diffusion image generation via linear latent decoding and curriculum learning, outperforming prior methods by over 30% F1 while reducing GPU memory by 97% and projection time to 0.2 seconds.
-
SurFITR: A Dataset for Surveillance Image Forgery Detection and Localisation
SurFITR is a new collection of 137k+ surveillance-style forged images that causes existing detectors to degrade while enabling substantial gains when used for training in both in-domain and cross-domain settings.
-
RefineAnything: Multimodal Region-Specific Refinement for Perfect Local Details
RefineAnything is a multimodal diffusion model using Focus-and-Refine crop-and-resize with blended paste-back to achieve high-fidelity local image refinement and near-perfect background preservation.
-
1.x-Distill: Breaking the Diversity, Quality, and Efficiency Barrier in Distribution Matching Distillation
1.x-Distill achieves better quality and diversity than prior few-step distillation methods at 1.67 and 1.74 effective NFEs on SD3 models with up to 33x speedup.
-
Banana100: Breaking NR-IQA Metrics by 100 Iterative Image Replications with Nano Banana Pro
Banana100 dataset shows that none of 21 popular NR-IQA metrics consistently rate images degraded by 100 iterative edits lower than clean originals.
-
Gen-Searcher: Reinforcing Agentic Search for Image Generation
Gen-Searcher is the first trained search-augmented image generation agent using SFT followed by GRPO reinforcement learning with dual text-image rewards, delivering 15-16 point gains on knowledge-intensive benchmarks.
-
DLEBench: Evaluating Small-scale Object Editing Ability for Instruction-based Image Editing Model
DLEBench is the first benchmark for small-scale object editing in instruction-based image editing models, using 1889 samples, seven instruction types, and a dual-mode evaluation protocol to reveal performance gaps in 10 tested models.
-
Beyond VLM-Based Rewards: Diffusion-Native Latent Reward Modeling
DiNa-LRM introduces a diffusion-native latent reward model using a noise-calibrated Thurstone likelihood on noisy states, matching VLM performance at lower compute in image alignment and preference optimization.
-
PlanViz: Evaluating Planning-Oriented Image Generation and Editing for Computer-Use Tasks
PlanViz is a new benchmark with three sub-tasks and PlanScore metric to evaluate planning-oriented image generation and editing by unified multimodal models for computer-use tasks.
-
VISTA-Bench: Do Vision-Language Models Really Understand Visualized Text as Well as Pure Text?
VISTA-Bench shows vision-language models degrade on visualized text in images compared to equivalent pure text, with larger gaps under increased perceptual difficulty.
-
AnomalyVFM -- Transforming Vision Foundation Models into Zero-Shot Anomaly Detectors
AnomalyVFM converts vision foundation models into zero-shot anomaly detectors via three-stage synthetic dataset generation plus low-rank adapters and weighted pixel loss, reaching 94.1% average image AUROC across nine datasets.
-
FreeOrbit4D: Training-Free Arbitrary Camera Redirection for Monocular Videos via Foreground-Complete 4D Reconstruction
FreeOrbit4D recovers a foreground-complete 4D proxy via decoupled background and object-centric reconstruction to provide geometric guidance for large-angle camera redirection in monocular videos using conditional video diffusion.
-
ColorConceptBench: A Benchmark for Probabilistic Color-Concept Understanding in Text-to-Image Models
ColorConceptBench shows leading text-to-image models lack sensitivity to abstract semantic color associations even with guidance scaling.
-
A Unified and Controllable Framework for Layered Image Generation with Visual Effects
LASAGNA produces layered images with integrated visual effects in a single pass, enabling drift-free edits via alpha compositing while releasing a 48K dataset and a 242-sample benchmark.
-
LooseRoPE: Content-aware Attention Manipulation for Semantic Harmonization
LooseRoPE modulates RoPE in diffusion attention maps to continuously trade off between preserving a pasted object's identity and harmonizing it with its new surroundings.
-
Beyond Patches: Global-aware Autoregressive Model for Multimodal Few-Shot Font Generation
GAR-Font is a global-aware autoregressive framework for multimodal few-shot font generation that adds global tokenization, a language-style adapter, and post-refinement to improve style coherence over patch-based methods.
-
InstructMoLE: Instruction-Guided Mixture of Low-rank Experts for Multi-Conditional Image Generation
InstructMoLE replaces per-token routing with instruction-guided global routing for mixture-of-low-rank-experts in diffusion transformers and adds an output-space orthogonality loss to improve multi-conditional image generation.