Edit2Perceive: Image Editing Diffusion Models Are Strong Dense Perceivers

TL;DR

Edit2Perceive leverages image editing diffusion models to achieve state-of-the-art dense perception across depth, normal, and matting tasks, emphasizing structure preservation and efficiency.

Contribution

The paper introduces a unified diffusion framework, Edit2Perceive, that adapts editing models for dense perception tasks with structure-preserving refinement and faster inference.

Findings

State-of-the-art results across depth, normal, and matting tasks

Effective structure-preserving refinement during denoising

Faster inference with single-step deterministic approach

Abstract

Recent advances in diffusion transformers have shown remarkable generalization in visual synthesis, yet most dense perception methods still rely on text-to-image (T2I) generators designed for stochastic generation. We revisit this paradigm and show that image editing diffusion models are inherently image-to-image consistent, providing a more suitable foundation for dense perception task. We introduce Edit2Perceive, a unified diffusion framework that adapts editing models for depth, normal, and matting. Built upon the FLUX.1 Kontext architecture, our approach employs full-parameter fine-tuning and a pixel-space consistency loss to enforce structure-preserving refinement across intermediate denoising states. Moreover, our single-step deterministic inference yields up to faster runtime while training on relatively small datasets. Extensive experiments demonstrate comprehensive state-of-the-art results across all three tasks, revealing the strong potential of editing-oriented diffusion transformers for geometry-aware perception.