REVIEW 3 cited by
Simple ReFlow: Improved Techniques for Fast Flow Models
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
Signed reviews
abstract
Diffusion and flow-matching models achieve remarkable generative performance but at the cost of many sampling steps, this slows inference and limits applicability to time-critical tasks. The ReFlow procedure can accelerate sampling by straightening generation trajectories. However, ReFlow is an iterative procedure, typically requiring training on simulated data, and results in reduced sample quality. To mitigate sample deterioration, we examine the design space of ReFlow and highlight potential pitfalls in prior heuristic practices. We then propose seven improvements for training dynamics, learning and inference, which are verified with thorough ablation studies on CIFAR10 $32 \times 32$, AFHQv2 $64 \times 64$, and FFHQ $64 \times 64$. Combining all our techniques, we achieve state-of-the-art FID scores (without / with guidance, resp.) for fast generation via neural ODEs: $2.23$ / $1.98$ on CIFAR10, $2.30$ / $1.91$ on AFHQv2, $2.84$ / $2.67$ on FFHQ, and $3.49$ / $1.74$ on ImageNet-64, all with merely $9$ neural function evaluations.
Forward citations
Cited by 3 Pith papers
-
IB-Flow: Information Bottleneck-Guided CFG Distillation for Few-Step Text-to-Image Generation
Information-Bottleneck closed forms for adaptive CFG supervisor timestep and strength yield SOTA 2-NFE text-to-image fidelity across FLUX, OpenUni, and Qwen-Image.
-
SnapGen-V: Generating a Five-Second Video within Five Seconds on a Mobile Device
SnapGen-V prunes, searches, and adversarially distills a video diffusion model down to 0.6B parameters that generates a five-second, 512x512 video on an iPhone 16 Pro Max in under five seconds.
-
Latent Schrodinger Bridge: Prompting Latent Diffusion for Fast Unpaired Image-to-Image Translation
Using prompt optimization and an SNR-matching change of variables, pretrained Stable Diffusion can approximate the source, target, and noise predictors of a Schrödinger bridge ODE, enabling 8-step unpaired image translation.
Discussion (0). Continue with ORCID to comment.