REVIEW 45 cited by
Adversarial Diffusion Distillation
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
read the original abstract
We introduce Adversarial Diffusion Distillation (ADD), a novel training approach that efficiently samples large-scale foundational image diffusion models in just 1-4 steps while maintaining high image quality. We use score distillation to leverage large-scale off-the-shelf image diffusion models as a teacher signal in combination with an adversarial loss to ensure high image fidelity even in the low-step regime of one or two sampling steps. Our analyses show that our model clearly outperforms existing few-step methods (GANs, Latent Consistency Models) in a single step and reaches the performance of state-of-the-art diffusion models (SDXL) in only four steps. ADD is the first method to unlock single-step, real-time image synthesis with foundation models. Code and weights available under https://github.com/Stability-AI/generative-models and https://huggingface.co/stabilityai/ .
Forward citations
Cited by 45 Pith papers
-
NEvo: Neural-Guided Evolutionary Video Synthesis for Dynamic Visual Selectivity
NEvo performs evolutionary search guided by a dynamic voxel-level encoding model to synthesize videos that maximize predicted activity in target brain ROIs, recovering known selectivities and revealing temporal dynami...
-
Language-Assisted Super-Resolution from Real-World Low-Resolution Patches
LA-SR redefines unpaired super-resolution in language space by projecting images into a semantically rich representation and applying vision-language model guided losses to handle real-world degradations extracted fro...
-
Oracle Noise: Faster Semantic Spherical Alignment for Interpretable Latent Optimization
Oracle Noise optimizes diffusion model noise on a Riemannian hypersphere guided by key prompt words to preserve the Gaussian prior, eliminate norm inflation, and achieve faster semantic alignment than Euclidean methods.
-
$Z^2$-Sampling: Zero-Cost Zigzag Trajectories for Semantic Alignment in Diffusion Models
Z²-Sampling implicitly realizes zero-cost zigzag trajectories for curvature-aware semantic alignment in diffusion models by reducing multi-step paths via operator dualities and temporal caching while synthesizing a di...
-
PromptEvolver: Prompt Inversion through Evolutionary Optimization in Natural-Language Space
PromptEvolver recovers high-fidelity natural language prompts for given images by evolving them via genetic algorithm guided by a vision-language model, outperforming prior methods on benchmarks.
-
It's Never Too Late: Noise Optimization for Collapse Recovery in Trained Diffusion Models
Noise optimization during sampling recovers diversity in mode-collapsed diffusion models while preserving output fidelity.
-
One Step Diffusion via Shortcut Models
Shortcut models enable high-quality single or few-step sampling in diffusion models with one network and training phase by conditioning on desired step size.
-
A Decomposable Probe for Few-Step Diffusion Models: Prompt, Latent, and Score Selectivity across Backbone Families and Distillation Paradigms
A three-layer perturbation probe shows latent selectivity is a near-binary rectified-flow fingerprint that survives ADD distillation, while score selectivity tracks distillation objective across 23 T2I models.
-
Diagnosing Under-Development of Irreversible Processes in Video Generation
Text-to-video generators under-develop irreversible processes: near-zero progress and 92-100% stasis versus real footage (rho=+0.40, 35% static), confirmed by human raters.
-
Parallel Decoding Distillation for Fast Image and Video Generation
A trajectory-based distillation method trains a student to predict multiple mean velocities per network evaluation, enabling 4-8 step generation with competitive quality and improved diversity.
-
DiTango: Cost-Effective Parallel Diffusion Generation with Selective Attention State Reuse
Selective reuse of composed attention states across denoising steps lets DiTango skip both remote KV communication and attention compute for low-contribution sequence partitions, cutting multi-GPU diffusion latency by...
-
DiffusionBench: On Holistic Evaluation of Diffusion Transformers
NanoGen unifies DiT training on ImageNet and T2I, reveals negative Pearson correlations (-0.377 to -0.580) in method rankings across metrics from 21 models, and motivates DiffusionBench for holistic evaluation.
-
Curvature-Adaptive Consistency Flow Matching: Autonomous Trajectory Optimization via Reinforcement Learning
CACFM applies RL to adaptively select critical regions in probability flow ODE trajectories for consistency distillation, yielding SOTA few-step results on FLUX and SDXL.
-
Addressing Detail Bottlenecks in Latent Diffusion for RGB-to-SWIR Image Translation
Introduces SCAE with skip connections and LGE to fix detail loss in LDMs for RGB-to-SWIR translation, yielding up to 2x mAP gains and 3.4x on small objects while reaching SOTA FID.
-
Variance Reduction for Expectations with Diffusion Teachers
CARV amortizes upstream diffusion teacher costs over noise resamples with timestep importance sampling and stratified-inverse-CDF sampling, delivering 2-3x effective compute gains in text-to-3D experiments and order-o...
-
Unlocking Complex Visual Generation via Closed-Loop Verified Reasoning
CLVR couples verified logical planning with pixel diffusion, uses proxy reinforcement learning on distilled histories, and merges weights to cut inference to 4 NFEs while outperforming open-source T2I models on comple...
-
Unlocking Complex Visual Generation via Closed-Loop Verified Reasoning
CLVR framework adds closed-loop visual verification, proxy prompt reinforcement learning, and delta-space weight merge to improve complex text-to-image generation over single-step or unverified multi-step baselines.
-
MetaSR: Content-Adaptive Metadata Orchestration for Generative Super-Resolution
MetaSR adaptively orchestrates metadata in a DiT-based generative SR model to deliver up to 1 dB PSNR gains and 50% bitrate savings across diverse content and degradations.
-
BiasIG: Benchmarking Multi-dimensional Social Biases in Text-to-Image Models
BiasIG is a multi-dimensional benchmark for social biases in T2I models that shows debiasing interventions frequently cause confounding discrimination effects.
-
Attention, May I Have Your Decision? Localizing Generative Choices in Diffusion Models
Implicit generative choices in diffusion models for ambiguous prompts are localized principally in self-attention layers, enabling a targeted ICM steering method that outperforms prior debiasing approaches.
-
Na\"ive PAINE: Lightweight Text-to-Image Generation Improvement with Prompt Evaluation
A lightweight predictor ranks initial noises by expected human-preference score for a prompt, selecting the best few for diffusion generation and reporting prompt difficulty.
-
Category-level Text-to-Image Retrieval Improved: Bridging the Domain Gap with Diffusion Models and Vision Encoders
Category-level retrieval improves when text queries are converted into multiple generated images, aggregated with a learned attention module, and fused with CLIP text similarity.
-
FLUX.1 Kontext: Flow Matching for In-Context Image Generation and Editing in Latent Space
FLUX.1 Kontext unifies image generation and editing via flow matching and sequence concatenation, delivering improved multi-turn consistency and speed on the new KontextBench benchmark.
-
Representation Alignment for Generation: Training Diffusion Transformers Is Easier Than You Think
Aligning noisy hidden states in diffusion transformers to clean features from pretrained visual encoders speeds up training over 17x and reaches FID 1.42.
-
Hunyuan-DiT: A Powerful Multi-Resolution Diffusion Transformer with Fine-Grained Chinese Understanding
Hunyuan-DiT is a new multi-resolution diffusion transformer that achieves state-of-the-art Chinese text-to-image generation through custom architecture, data pipelines, and multimodal caption refinement.
-
Scaling Rectified Flow Transformers for High-Resolution Image Synthesis
Biased noise sampling for rectified flows combined with a bidirectional text-image transformer architecture yields state-of-the-art high-resolution text-to-image results that scale predictably with model size.
-
DiFaReli++: Diffusion Face Relighting with Consistent Cast Shadows
DiFaReli++ conditions a DDIM on shading references and inferred shadow maps to relight single-view faces with consistent shadows, trained only on 2D images and claiming SOTA on Multi-PIE.
-
Examining the Efficacy of Graph Neural Network Message-Passing in Regression Contexts
Across four NAS/DNN-predictor regression benchmarks, GEN (deep graph convolution) achieves the best average rank over 11 GNN message-passing layers, though attention GATv2 wins on the largest graphs.
-
MicroZoom: Structure-Preserving Detail Synthesis at Extreme Scale
A cascaded, segmentation-conditioned, per-instance diffusion method synthesizes globally coherent gigapixel microscopic detail from a phone photo and sparse microscope references at up to 350×.
-
Mage-Flow: An Efficient Native-Resolution Foundation Model for Image Generation and Editing
A compact 4B image generation/editing system with a fast one-step VAE, native-resolution packing, RL alignment, and 4-step distillation reports competitive benchmarks against 6B–80B open models.
-
Language-Assisted Super-Resolution from Real-World Low-Resolution Patches
LA-SR extracts real LR patches from depth-varying regions in single images and uses vision-language models with linguistic content and quality losses for unpaired super-resolution.
-
Scenario-conditioned flow matching for probabilistic generation of three-component ground-motion waveforms
WaveFlowGMM generates scenario-conditioned three-component ground-motion waveforms by using symbolic learning for PGA amplitude and AlphaFlow for normalized wavelet-packet waveforms that are later rescaled.
-
Curvature-Adaptive Consistency Flow Matching: Autonomous Trajectory Optimization via Reinforcement Learning
A reinforcement-learning agent that picks high-error trajectory segments during consistency distillation improves few-step text-to-image generation on FLUX and SDXL.
-
On the Redundancy of Timestep Embeddings in Diffusion Models
Under high-dimensional concentration conditions, the diffusion denoising objective admits the same global minimizer without timestep embeddings, and time-agnostic U-Nets/DiTs empirically match or improve FID on CelebA...
-
EPIG: Emotion-Based Prompting for Personalised Image Generation
EPIG is a training-free prompt enrichment technique using valence-arousal representations that reduces mean arousal error by 14% versus naive insertion and 12% versus LLM expansion on a 10-prompt benchmark while prese...
-
LUCID: Learning Unified Control for Image Deflaring and Exposure Mastery in Nighttime Photography
LUCID introduces a unified controllable framework for nighttime image restoration that disentangles flares and uses diffusion priors with four-mode training for selective exposure and artifact control.
-
Lens: Rethinking Training Efficiency for Foundational Text-to-Image Models
Lens is a 3.8B-parameter text-to-image model that reaches competitive or superior performance to >6B-parameter systems using 19.3% of the training compute of Z-Image through a densely captioned 800M dataset, multi-res...
-
Variance Reduction for Expectations with Diffusion Teachers
CARV introduces a hierarchical Monte Carlo estimator with amortized reuse, importance sampling, and stratification that yields 2-3x effective compute gains on diffusion-teacher pipelines while cutting gradient varianc...
-
MONET: A Massive, Open, Non-redundant and Enriched Text-to-image dataset
MONET is an open 104.9M image-text pair dataset created via safety filtering, deduplication, and multi-VLM recaptioning from 2.9B raw pairs, validated by training a competitive 4B-parameter latent diffusion model.
-
Brain-to-Image Retrieval and Reconstruction via Multimodal EEG Alignment
A multimodal alignment pipeline decodes EEG signals recorded during natural image viewing into image retrieval (86.3% Top-1) and reconstruction (CLIP 0.903) tasks.
-
Attention, May I Have Your Decision? Localizing Generative Choices in Diffusion Models
Implicit generative choices in diffusion models concentrate in self-attention layers; targeted ICM interventions there outperform broader debiasing methods with fewer artifacts.
-
Systematic Optimization of Real-Time Diffusion Model Inference on Apple M3 Ultra
Systematic benchmarking of diffusion model optimizations on Apple M3 Ultra produces 22.7 FPS real-time img2img at 512x512 and demonstrates that CUDA-derived techniques do not transfer directly to Apple Silicon.
-
On the Redundancy of Timestep Embeddings in Diffusion Models
Timestep embeddings are redundant in diffusion models under certain conditions, with time-agnostic variants matching or exceeding conditioned models on FID, precision, and recall for CelebA and CIFAR-10.
-
Learning from Limited and Imperfect Data
A doctoral thesis compiling nine peer-reviewed papers on long-tailed image generation, long-tailed recognition, semi-supervised learning, and domain adaptation.
-
Sora: A Review on Background, Technology, Limitations, and Opportunities of Large Vision Models
The paper reviews the background, technology, applications, limitations, and future directions of OpenAI's Sora text-to-video generative model based on public information.
Discussion (0). Sign in to comment.