DIAMOND: Directed Inference for Artifact Mitigation in Flow Matching Models

TL;DR

DIAMOND is a training-free inference method that corrects generative trajectories in diffusion models to reduce artifacts, enabling high-fidelity, artifact-free image synthesis without additional training or model modifications.

Contribution

The paper introduces DIAMOND, a novel trajectory correction technique that mitigates artifacts during inference in diffusion models without requiring training or weight changes.

Findings

Effective artifact reduction during inference.

Applicable to standard diffusion models without retraining.

Produces high-quality, artifact-free images in zero-shot settings.

Abstract

Despite impressive results from recent text-to-image models like FLUX, visual and anatomical artifacts remain a significant hurdle for practical and professional use. Existing methods for artifact reduction, typically work in a post-hoc manner, consequently failing to intervene effectively during the core image formation process. Notably, current techniques require problematic and invasive modifications to the model weights, or depend on a computationally expensive and time-consuming process of regional refinement. To address these limitations, we propose DIAMOND, a training-free method that applies trajectory correction to mitigate artifacts during inference. By reconstructing an estimate of the clean sample at every step of the generative trajectory, DIAMOND actively steers the generation process away from latent states that lead to artifacts. Furthermore, we extend the proposed method to standard Diffusion Models, demonstrating that DIAMOND provides a robust, zero-shot path to high-fidelity, artifact-free image synthesis without the need for additional training or weight modifications in modern generative architectures. Code is available at https://gmum.github.io/DIAMOND/