On-device research index

arXiv research

A locally-built, LLM-digested index of recent arXiv papers in quant finance, geometry/topology, and statistical ML — keyword search served straight from SQLite on this machine.

168,657 papers · 148 categories

Trend · papers per month

count % of corpus

15 results for “FFHQ”

Towards More General Control of Diffusion Models Using Jeffrey Guidancecs.LG

Jeffrey guidance extends diffusion-model control to more complex applications.

problem Controlling diffusion models beyond simple cases like conditional sampling.

method Leveraging Jeffrey's rule of conditioning to update marginal distributions towards a target distribution.

result Significant reductions in FID on CIFAR-10 and FFHQ with Inception embeddings as the target.

Jeffrey guidance diffusion models FID CIFAR-10 FFHQ

2026-06-11abs ↗pdf ↗

Exact Posterior Score Estimation for Solving Linear Inverse Problemscs.LG

Exact posterior score estimation for solving linear inverse problems

problem Solving linear inverse problems

method Derive the exact posterior score and use it as a denoising training objective

result EPS outperforms training-free and training-based baselines on various metrics

linear inverse posterior score denoising FFHQ ImageNet

2026-06-15abs ↗pdf ↗

Improved Techniques for Training Score-Based Generative Modelscs.LG

Improved score-based models generate high-quality images up to 256x256.

problem Training score-based models for high-resolution images is unstable and limited.

method Theoretical analysis, exponential moving average of model weights.

result Score-based models can generate high-fidelity images up to 256x256.

score-based models generative models image generation exponential moving average high-resolution

2020-06-16abs ↗pdf ↗

Improving Decoupled Posterior Sampling for Inverse Problems using Data Consistency Constraintcs.LG

Improved diffusion models for inverse problems by integrating data consistency constraints.

problem Errors in earlier steps of diffusion models during posterior sampling.

method Guided Decoupled Posterior Sampling (GDPS) with data consistency constraint.

result GDPS achieves state-of-the-art performance, improving accuracy over existing methods.

inverse problems diffusion models posterior sampling data consistency latent diffusion

2024-12-01abs ↗pdf ↗

Disentangled Inference for GANs with Latently Invertible Autoencodercs.LG

Generative Adversarial Networks (GANs) play an increasingly important role in machine learning. However, there is one fundamental issue hindering their practical applications: the absence of capability for encoding real-world samples. The conventional way of addressing this issue is to learn an encoder for GAN via Vari…

2019-06-19abs ↗pdf ↗

Toward Architecture-Agnostic Local Control of Posterior Collapse in VAEscs.LG

New method controls posterior collapse in VAEs without network architecture constraints.

problem Posterior collapse in VAEs reduces diversity of generated samples.

method Introduces Latent Reconstruction (LR) loss to control posterior collapse.

result Controls posterior collapse on various datasets without architectural constraints.

VAE posterior collapse local posterior collapse Latent Reconstruction (LR) loss mathematical properties

2025-08-17abs ↗pdf ↗

When Are Two Scores Better Than One? Investigating Ensembles of Diffusion Modelscs.LG

This study examines ensembling of diffusion models for improved generative quality.

problem Improving generative quality with ensembling of score-based diffusion models.

method Investigated ensembling of scores from multiple diffusion models on image and tabular data.

result Ensembling scores generally improves model likelihood and score-matching loss but not perceptual quality metrics.

diffusion models ensemble methods score-based models generative modeling image quality

2026-01-16abs ↗pdf ↗

LightSBB-M: Bridging Schrödinger and Bass for Generative Diffusion Modelingcs.LG

LightSBB-M improves generative diffusion modeling with lower 2-Wasserstein distances.

problem Improving generative diffusion models using Schrödinger Bridge and Bass methods.

method Optimizes SBB transport plan with dual representation and tunable beta parameter.

result Achieves up to 32% improvement in 2-Wasserstein distance on synthetic datasets.

Generative Diffusion Modeling Schrödinger Bridge Bass Martingale 2-Wasserstein Distance Image-to-Image Translation

2026-01-27abs ↗pdf ↗

On the Stability of Iterative Retraining of Generative Models on their own Datacs.LG

The paper studies stability of generative models trained on mixed data.

problem Training generative models on mixed datasets (real and synthetic data).

method Developed a framework to rigorously study stability under specific conditions.

result Proved the stability of iterative training under certain conditions.

generative models mixed datasets stability iterative training self-consuming models

2023-09-30abs ↗pdf ↗

Are First-Order Diffusion Samplers Really Slower? A Fast Forward-Value Approachstat.ML

A new first-order sampler improves diffusion probabilistic model sampling quality.

problem The belief that first-order methods are inherently slower for diffusion probabilistic model sampling.

method A novel training-free, first-order sampler that approximates the forward-value evaluation via a one-step lookahead predictor.

result The proposed sampler provably approximates the ideal forward-value trajectory while retaining first-order convergence and can improve sample quality under the same NFE budget.

diffusion sampling first-order methods forward-value evaluation lookahead predictor sampling quality

2025-12-31abs ↗pdf ↗

NVAE: A Deep Hierarchical Variational Autoencoderstat.ML

NVAE improves VAE performance on large image datasets.

problem Improving variational autoencoder performance for large image datasets.

method Deep hierarchical VAE with depth-wise separable convolutions and batch normalization, residual parameterization of Normal distributions, and spectral regularization.

result NVAE achieves state-of-the-art results on MNIST, CIFAR-10, CelebA 64, and CelebA HQ datasets.

VAE Hierarchical Normalizing Flows Image Generation Spectral Regularization

2020-07-08abs ↗pdf ↗

Beyond First-Order Tweedie: Solving Inverse Problems using Latent Diffusioncs.LG

Paper introduces STSL, a second-order Tweedie sampler for efficient posterior sampling in inverse problems.

problem Computational challenges in sampling from posterior distributions using latent diffusion models.

method Introduces STSL, a novel second-order Tweedie sampler with tractable reverse process.

result STSL achieves 4X and 8X reduction in neural function evaluations compared to state-of-the-art solvers.

latent diffusion inverse problems posterior sampling second-order approximation neural function evaluations

2023-12-01abs ↗pdf ↗

Variational Bayes image restoration with compressive autoencoderscs.CV

A new method uses compressive autoencoders for image restoration.

problem Efficient regularization of inverse problems in computational imaging.

method Variational Bayes Latent Estimation (VBLE) with compressive autoencoders.

result VBLE achieves similar performance to state-of-the-art PnP methods but faster.

image restoration variational Bayes compressive autoencoders plug-and-play latent estimation

2023-11-29abs ↗pdf ↗

Multimeasurement Generative Modelsstat.ML

Generative models learn smoother densities to sample from unknown distributions.

problem Sampling from unknown distributions in high-dimensional spaces.

method Formalizes sampling problem, introduces multimeasurement noise model, derives Bayes estimator, and uses underdamped Langevin MCMC.

result Formulation leads to efficient sampling methods and theoretical connections with denoising autoencoders.

generative models sampling Bayes estimator multimeasurement noise model underdamped Langevin MCMC

2021-12-18abs ↗pdf ↗

Blur, Noise, and Compression Robust Generative Adversarial Networkscs.CV

BNCR-GAN improves GANs to generate clean images from degraded inputs.

problem Generating clean images from blurred, noisy, and compressed degraded inputs.

method Multiple-generator model with image, blur-kernel, noise, and quality-factor generators, using masking architectures and adaptive consistency losses.

result BNCR-GAN effectively learns clean image generators from degraded images without degradation parameters.

Generative Adversarial Networks Blur Noise Compression Robustness

2020-03-17abs ↗pdf ↗