Preserve Anything: Controllable Image Synthesis with Object Preservation

TL;DR

This paper presents 'Preserve Anything', a new controllable image synthesis method that enhances object preservation, semantic consistency, and user control in text-to-image generation, achieving state-of-the-art results and a new benchmark dataset.

Contribution

It introduces a novel framework with object preservation and background guidance modules, along with a comprehensive benchmark dataset for evaluating image synthesis quality.

Findings

Achieves state-of-the-art FID 15.26 and CLIP-S 32.85 scores.

Significantly improves prompt alignment and photorealism in user studies.

Provides a new dataset with 240K natural and 18K synthetic images for evaluation.

Abstract

We introduce \textit{Preserve Anything}, a novel method for controlled image synthesis that addresses key limitations in object preservation and semantic consistency in text-to-image (T2I) generation. Existing approaches often fail (i) to preserve multiple objects with fidelity, (ii) maintain semantic alignment with prompts, or (iii) provide explicit control over scene composition. To overcome these challenges, the proposed method employs an N-channel ControlNet that integrates (i) object preservation with size and placement agnosticism, color and detail retention, and artifact elimination, (ii) high-resolution, semantically consistent backgrounds with accurate shadows, lighting, and prompt adherence, and (iii) explicit user control over background layouts and lighting conditions. Key components of our framework include object preservation and background guidance modules, enforcing lighting consistency and a high-frequency overlay module to retain fine details while mitigating unwanted artifacts. We introduce a benchmark dataset consisting of 240K natural images filtered for aesthetic quality and 18K 3D-rendered synthetic images with metadata such as lighting, camera angles, and object relationships. This dataset addresses the deficiencies of existing benchmarks and allows a complete evaluation. Empirical results demonstrate that our method achieves state-of-the-art performance, significantly improving feature-space fidelity (FID 15.26) and semantic alignment (CLIP-S 32.85) while maintaining competitive aesthetic quality. We also conducted a user study to demonstrate the efficacy of the proposed work on unseen benchmark and observed a remarkable improvement of $\sim25\%$, $\sim19\%$, $\sim13\%$, and $\sim14\%$ in terms of prompt alignment, photorealism, the presence of AI artifacts, and natural aesthetics over existing works.