This work improves diffusion models by estimating the optimal loss value for better training diagnostics.
problem The optimal loss value of diffusion models is unknown and not indicative of absolute data-fitting quality.
method Derive the optimal loss in closed form and develop effective estimators, including a stochastic variant.
result Unlocking the optimal loss as a metric for diagnosing training quality of diffusion models.
Improved diffusion bridge sampling with rKL-LD loss.
problem Improving sampling from unnormalized distributions using diffusion bridges.
method Employing the rKL-LD loss instead of the Log Variance (LV) loss for diffusion bridges.
result rKL-LD consistently outperforms LV loss in diffusion bridges.
New reweighted losses improve diffusion model training and image quality.
problem Training and improving diffusion models for image generation.
method Constructing a cascade of time-dependent variational lower bounds.
result Significant improvements in pixel-space image modeling quality.
Zero-shot contrastive loss improves text-guided image style transfer without extra training.
problem Stochastic nature of diffusion models leads to trade-offs between style transformation and content preservation.
method Proposes a zero-shot contrastive loss for diffusion models that doesn't require additional fine-tuning or auxiliary networks.
result Method outperforms existing methods while preserving content and requiring no additional training.
The paper justifies time-dependent loss reweighting schemes for flow matching and diffusion models.
problem Theoretical justification for time-dependent loss reweighting schemes in flow matching and diffusion models.
method Clarifies that the loss can depend on both time and state, and shows theoretical justification for time-dependent loss weighting schemes.
result Time-dependent loss weighting schemes are theoretically justified for Generator Matching and Edit Flows.
Improved speech enhancement using diffusion models with MSE loss.
problem Efficient incorporation of noisy speech in generative speech enhancement.
method Augmented diffusion-based generative model with a MSE loss for enhanced speech.
result Proposed method improves speech enhancement performance compared to original diffusion model.
New loss function improves accuracy of MRI parameter estimation.
problem Systematic errors in parameter estimates at low SNR.
method Developed and implemented negative log Rician likelihood (NLR) loss.
result NLR loss shows higher accuracy in parameter estimation than MSE loss at low SNR.
A new method for image translation using disentangled style and content preservation.
problem Difficulty in maintaining original content during reverse diffusion in diffusion-based image translation.
method Disentangled style and content representation using intermediate keys from ViT model, CLIP loss, semantic divergence loss, and resampling strategy.
result Outperforms state-of-the-art models in text-guided and image-guided translation tasks.
Improved sampling via learned diffusions using variational losses.
problem Sampling from target distributions without direct access to samples.
method Generalized Schrödinger bridge problem, variational formulation, gradient-based optimization.
result Proposed log-variance loss leads to improved performance.
DiffEnc improves diffusion models by adding flexibility and achieving better likelihood on CIFAR-10.
problem Improving the likelihood of diffusion models on image datasets.
method Introducing a data- and depth-dependent mean function and a free weight parameter for noise variance.
result Achieved statistically significant improvement in likelihood on CIFAR-10.
Soft Truncation improves diffusion model performance by balancing loss scales across diffusion times.
problem Inverse correlation between density estimation and sample generation in diffusion models.
method Introduces Soft Truncation, a training technique that softens the truncation hyperparameter.
result Soft Truncation achieves state-of-the-art performance on various datasets.
DreamFusion uses text-to-image diffusion models to create 3D images efficiently.
problem Lack of large-scale 3D datasets and efficient architectures for 3D synthesis.
method Adapting a 2D diffusion model to 3D synthesis using a loss based on probability density distillation.
result A 3D model can be optimized from a 2D diffusion model, allowing for text-to-3D synthesis.
Paper proposes a method to speed up discrete diffusion models by distilling many steps into few.
problem Challenges in capturing dependencies between elements in discrete diffusion models.
method Proposes 'mixture' models and loss functions to distill many sampling steps into few.
result Effective in distilling pretrained discrete diffusion models across image and language domains.
Maximum likelihood training improves the performance of score-based diffusion models.
problem Training score-based diffusion models with maximum likelihood.
method Trained by minimizing a weighted combination of score matching losses, with a specific weighting scheme that bounds negative log-likelihood.
result Maximum likelihood training improves the log-likelihood of score-based diffusion models across multiple datasets.
This paper solves hedging in incomplete markets using neural networks.
problem Hedging in incomplete markets with risk factor, illiquidity, and discrete transaction dates.
method Proposes a jump-diffusion model and uses RNN, LSTM, and Mogrifier-LSTM neural networks for hedging strategies.
result Mogrifier-LSTM is the fastest and most effective model for hedging.
Pixel-space diffusion models outperform latent models on high-resolution image synthesis.
problem Efficiency and quality trade-off in high-resolution image synthesis.
method Sigmoid loss-weighting, simplified architecture, and resolution scaling.
result Achieved 1.5 FID on ImageNet512, new SOTA results on other datasets.
Diffusion-QL uses diffusion models to improve offline RL performance.
problem Offline RL struggles with function approximation errors on out-of-distribution actions.
method Diffusion-QL represents the policy as a conditional diffusion model and optimizes action-values.
result Diffusion-QL achieves state-of-the-art performance on D4RL benchmark tasks.
Improves generative models by adding jump-diffusion noise.
problem Limited performance of diffusion models in generating samples from unknown distributions.
method Generalizes diffusion processes to include jump-diffusion noise, deriving closed-form generalized score functions.
result Jump-diffusion models outperform Gaussian models in specific parameter regimes.
Uniform diffusion approximation for SGD in non-convex settings.
problem Finite-time diffusion approximation for SGD.
method Establishing uniform-in-time diffusion approximation with strong convexity and mild conditions.
result Uniform-in-time diffusion approximation of SGD without convexity of each loss function.
Study evaluates ensemble methods for zero-shot uncertainty quantification with diffusion models.
problem Quantifying uncertainty in zero-shot regression problems using diffusion models.
method Used diffusion probabilistic models for ensemble prediction and evaluated their effectiveness on various regression tasks.
result Ensemble methods consistently improve model prediction accuracy across different regression tasks.
PRISMA uses PDE residuals for fast, robust, and accurate inference.
problem Slow gradient-based optimization and instability in PDE residual-based methods.
method Integrates PDE residuals directly into the model's architecture via attention mechanisms in the spectral domain.
result Competitive accuracy with significantly lower inference costs and faster speeds.
PDDS samples from unnormalized densities using iterative particle scheme.
problem Sampling from unnormalized probability densities.
method Iterative particle scheme with novel score matching loss.
result Asymptotically consistent estimates for multimodal and high-dimensional tasks.
Diffusion models accurately recover mixture weights from generated samples despite score function insensitivity.
problem Score-based generative models often fail to learn correct relative mode amplitudes (mixture weights) from generated samples.
method Relate diffusion score matching (DSM) loss to mixture weight estimation error, define diffusion score sensitivity index (DSSI), and prove its governing role in mixture weight recovery.
result Generated samples can accurately recover mixture weights from the DSM loss, even when the target score is insensitive to mixture weights.
The study uncovers the conditions under which diffusion models memorize or generalize.
problem Understanding the balance between memorization and generalization in diffusion models.
method Theoretical and mathematical framework to investigate memorization and generalization in diffusion models.
result Theoretical crossover point predicts a phase transition in diffusion models, validating the hypothesis.
ED-NeRF efficiently edits 3D scenes using latent space NeRF and improved loss functions.
problem Slow training speeds and inadequate editing loss functions in existing NeRF editing techniques.
method Embedding real-world scenes into latent space of LDM, using a unique refinement layer and an improved loss function.
result ED-NeRF achieves faster editing speed and improved output quality compared to state-of-the-art models.
New method learns diffusion bridges for rare events.
problem Simulating rare events in diffusion processes.
method Iterative online learning based on self-consistency.
result Strong performance in various empirical settings.
New method controls classifier guidance in diffusion models.
problem Improving classifier guidance in diffusion models.
method Cross-entropy control of classifier gradients.
result Effective guidance vectors with mean squared error O(dε). Replacing MSE with f-divergence in diffusion models improves robustness under data contamination.
problem Improving robustness of diffusion models under data contamination.
method Replacing MSE with f-divergence in diffusion models.
result Empirical improvement in performance under data contamination.
Anomalous diffusion in SGD reveals interactions between hyperparameters and Hessian.
problem Understanding the limiting dynamics of SGD in deep neural networks.
method Continuous-time model of SGD as an underdamped Langevin equation, derived for linear regression.
result Anomalous diffusion is explained by modified loss and probability currents in phase space.
Study on privacy leakage in noisy gradient descent algorithms.
problem Information leakage of iterative randomized learning algorithms about training data.
method Analyzes the dynamics of Rényi differential privacy loss in noisy gradient descent algorithms.
result Privacy loss converges exponentially fast for smooth and strongly convex loss functions.
Diffusion models achieve nearly optimal distribution estimation in various spaces.
problem Theoretical limitations of diffusion modeling for distribution estimation.
method Analysis of approximation and generalization abilities of diffusion models in Besov spaces.
result Diffusion models achieve nearly minimax optimal estimation rates in total variation and Wasserstein distances.
New score matching method estimates local intrinsic dimension efficiently.
problem Quantifying the local intrinsic dimension of complex data.
method Denoising score matching loss and equivalent implicit score matching loss.
result Denoising score matching loss is a highly competitive and scalable LID estimator.
Deep learning dynamics exhibit anomalous superdiffusion initially, aiding escape from local minima.
problem Understanding the dynamics of learning in deep neural networks.
method Novel analysis of SGD dynamics and loss landscape structure.
result SGD exhibits anomalous superdiffusion initially, transitioning to subdiffusion as learning progresses.
New method modifies diffusions for singular rewards.
problem Handling singular rewards in diffusions.
method Malliavin calculus for non-differentiable rewards.
result Stable and reliable training of diffusions.
Beta diffusion generates bounded data using multiplicative transitions.
problem Generating data within specific ranges.
method Integrates demasking and denoising with scaled and shifted beta distributions.
result KLUBs are more effective for optimizing beta diffusion compared to negative ELBOs.
DPS uses PINNs to estimate drift in diffusion models for sampling.
problem Accurately estimating drift term in reverse SDE from unnormalized density.
method Diffusion-PINN Sampler (DPS) solves PINN for log-density of SDE marginals.
result DPS achieves convergence guarantees and accurately samples complex distributions.
Single-step samplers generate high-quality samples efficiently.
problem Sampling from unnormalized distributions is computationally expensive.
method Developed consistent diffusion samplers that generate samples in a single step.
result Single-step samplers produce high-fidelity samples with less than 1% of traditional samplers' evaluations.
A new diffusion model tackles brightness issues with a probabilistic approach.
problem Brightness-related limitations in diffusion models.
method Introduces a novel diffusion model with a probabilistic framework, modifying both forward and reverse diffusion processes.
result The model mitigates brightness-related limitations and improves performance in high-dimensional settings.
New bounds close the score matching gap for diffusion models.
problem The difference between sample quality and score matching loss in diffusion models.
method Theoretical analysis of score matching gap, developing tighter bounds for KL divergence, reverse KL divergence, and Wasserstein distance.
result The quality of score approximation impacts closing the score matching gap for low noise scales.
This work analyzes PINNs for advection-diffusion equations using NTK theory.
problem Understanding and resolving the training difficulties of PINNs for advection-diffusion equations.
method Neural Tangent Kernel (NTK) analysis of PINNs for the linear advection-diffusion equation (LAD).
result PINNs struggle due to spectral bias and convergence rate disparity, especially in advection-dominated and diffusion-dominated regimes.
We propose to use Gaussian process regression to accurately estimate the diffusion MRI signal at arbitrary locations in q-space. By estimating the signal on a grid, we can do synthetic diffusion spectrum imaging: reconstructing the ensemble averaged propagator (EAP) by an inverse Fourier transform. We also propose an a…
Studying SGD on deep neural networks using diffusion maps.
problem Understanding why SGD performs well in deep learning.
method Data-driven approach using diffusion maps to analyze SGD dynamics.
result SGD dynamics may mainly live on a low-dimensional manifold in high-dimensional parameter space.
DEUA detects diffusion-generated images by accounting for different types of uncertainty.
problem Detecting generated images with varying aleatoric and epistemic uncertainty.
method DEUA framework using Laplace approximation for DEU estimation and asymmetric loss function.
result DEUA achieves state-of-the-art performance on large-scale benchmarks.
In vivo diffusion tensor imaging (DTI) is a promising technique to investigate noninvasively the fiber structures of the in vivo human heart. However, signal loss due to motions remains a persistent problem in in vivo cardiac DTI. We propose a novel motion-compensation method for investigating in vivo myocardium struct…
Study improves data quality assessment for structural monitoring data.
problem Ensuring reliability of structural health monitoring data.
method Probabilistic data quality assessment using a conditional diffusion model.
result Significantly improves accuracy of data quality assessment.
Simplified masked diffusion models improve discrete data generation.
problem Complex model formulations and unclear relationships hinder discrete data generative modeling.
method Developed a simple and general framework for masked diffusion models.
result Models trained on OpenWebText surpass prior diffusion language models and outperform autoregressive models.
We consider a diffusion approximation to an insurance risk model where an external driver models a stochastic environment. The insurer can buy reinsurance. Moreover, investment in a financial market is possible. The financial market is also driven by the environmental process. Our goal is to maximise terminal expected …
Unified approach combining BSDEs and PINNs for solving PDEs.
problem Solving high-dimensional partial differential equations.
method Interpolating between BSDEs and PINNs using diffusion loss.
result Unified understanding of numerical approaches for high-dimensional PDEs.