REVIEW 6 cited by
ReNO: Enhancing One-step Text-to-Image Models through Reward-based Noise Optimization
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
Text-to-Image (T2I) models have made significant advancements in recent years, but they still struggle to accurately capture intricate details specified in complex compositional prompts. While fine-tuning T2I models with reward objectives has shown promise, it suffers from "reward hacking" and may not generalize well to unseen prompt distributions. In this work, we propose Reward-based Noise Optimization (ReNO), a novel approach that enhances T2I models at inference by optimizing the initial noise based on the signal from one or multiple human preference reward models. Remarkably, solving this optimization problem with gradient ascent for 50 iterations yields impressive results on four different one-step models across two competitive benchmarks, T2I-CompBench and GenEval. Within a computational budget of 20-50 seconds, ReNO-enhanced one-step models consistently surpass the performance of all current open-source Text-to-Image models. Extensive user studies demonstrate that our model is preferred nearly twice as often compared to the popular SDXL model and is on par with the proprietary Stable Diffusion 3 with 8B parameters. Moreover, given the same computational resources, a ReNO-optimized one-step model outperforms widely-used open-source models such as SDXL and PixArt-$\alpha$, highlighting the efficiency and effectiveness of ReNO in enhancing T2I model performance at inference time. Code is available at https://github.com/ExplainableML/ReNO.
Forward citations
Cited by 6 Pith papers
-
Not All Parameters Matter: Masking Diffusion Models for Enhancing Generation Ability
MaskUNet masks U-Net weights with a timestep- and sample-dependent binary mask, improving zero-shot FID on COCO by about 1.1 to 1.5 points while leaving pre-trained weights frozen.
-
A Noise is Worth Diffusion Guidance
A one-step learned noise refinement replaces classifier-free guidance at inference on Stable Diffusion 2.1, giving comparable image quality at about 1.7x lower cost.
-
DyMO: Training-Free Diffusion Model Alignment with Dynamic Multi-Objective Scheduling
DyMO improves text-to-image outputs at inference time by dynamically scheduling an LLM-built semantic attention objective with a human-preference reward, without retraining the diffusion model.
-
CoCoNO: Attention Contrast-and-Complete for Initial Noise Optimization in Text-to-Image Synthesis
CoCoNO improves prompt-image alignment by optimizing the initial latent with attention complete and contrast losses, reducing subject neglect and mixed-subject artifacts.
-
TKG-DM: Training-free Chroma Key Content Generation Diffusion Model
Adjusting the mean of specific channels in the initial noise of Stable Diffusion produces foreground objects on a uniform, user-selected chroma key background without any fine-tuning.
-
Fast Direct: Query-Efficient Online Black-box Guidance for Diffusion-model Target Generation
Fast Direct guides diffusion models toward a black-box objective by repeatedly nudging the noise sequence toward a pseudo-target built from a Gaussian-process surrogate, claiming 6x to 44x query-efficiency gains over ...
Discussion (0). Continue with ORCID to comment.