Generative depth estimation methods leverage the rich visual priors stored in pre-trained text-to-image diffusion models, demonstrating astonishing zero-shot capability. However, parameter updates during training lead to catastrophic degradation in the image generation capability of the pre-trained model. We introduce MERGE, a unified model for image generation and depth estimation, starting from a fixed pre-trained text-to-image model. MERGE demonstrates that the pre-trained text-to-image model can do more than image generation, but also expand to depth estimation effortlessly. Specifically, MERGE introduces a play-and-plug framework that enables seamless switching between image generation and depth estimation modes through simple and pluggable converters. Meanwhile, we propose a Group Reuse Mechanism to encourage parameter reuse and improve the utilization of the additional learnable parameters. MERGE unleashes the powerful depth estimation capability of the pre-trained text-to-image model while preserving its original image generation ability. Compared to other unified models for image generation and depth estimation, MERGE achieves state-of-the-art performance across multiple depth estimation benchmarks. The code will be made available at https://github.com/H-EmbodVis/MERGE
@article{arxiv.2510.23574,
title = {More Than Generation: Unifying Generation and Depth Estimation via Text-to-Image Diffusion Models},
author = {Hongkai Lin and Dingkang Liang and Mingyang Du and Xin Zhou and Xiang Bai},
journal= {arXiv preprint arXiv:2510.23574},
year = {2025}
}
Comments
Accepted by NeurIPS 2025. The code will be made available at https://github.com/H-EmbodVis/MERGE