UniDepth: Universal Monocular Metric Depth Estimation

By Luigi Piccinelli, Yung-Hsu Yang, Christos Sakaridis, Mattia Segu, Siyuan Li, Luc Van Gool, Fisher Yu. Computer Vision and Pattern Recognition, 2024

455 citations recorded by the source corpus.

Accurate monocular metric depth estimation (MMDE) is crucial to solving downstream tasks in 3D perception and modeling. However, the remarkable accuracy of recent MMDE methods is confined to their training domains. These methods fail to generalize to unseen domains even in the presence of moderate domain gaps, which hinders their practical applicability. We propose a new model, UniDepth, capable of reconstructing metric 3D scenes from solely single images across domains. Departing from the existing MMDE methods, UniDepth directly predicts metric 3D points from the input image at inference time without any additional information, striving for a universal and flexible MMDE solution. In particular, UniDepth implements a self-promptable camera module predicting dense camera representation to condition depth features. Our model exploits a pseudo-spherical output representation, which disentangles camera and depth representations. In addition, we propose a geometric invariance loss that promotes the invariance of camera-prompted depth features. Thorough evaluations on ten datasets in a zero-shot regime consistently demonstrate the superior performance of UniDepth, even when compared with methods directly trained on the testing domains. Code and models are available at: github.com/lpiccinelli-eth/unidepth.

Subfields: Aerospace Engineering, Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/60bdb4c44136fc452170cc65fbd9cd8ad73ca425

Citations in the corpus

• Depth Anything V2 (2024, unknown): “Therefore, there have been numerous MDE models [56, 7, 6, 95, 26, 38, 31, 89, 88, 25, 20, 52, 28] emerging recently, which are all capable of addressing open-world images.”
• Depth Anything 3: Recovering the Visual Space from Any Views (2025, unknown): “We compare with state-of-the-art metric depth estimation methods, including DepthPro [7], Metric3D v2 [37], UniDepthv1 [65] and UniDepthv2 [66], on 5 benchmarks: NYUv2 [76], KITTI [25], ETH3D [72], . . . SUN-RGBD [81] and DIODE (indoor) [88].”
• Depth Pro: Sharp Monocular Metric Depth in Less Than a Second (2024, unknown): “Our baselines include Depth Anything (Yang et al., 2024a), Metric3D (Yin et al., 2023), PatchFusion (Li et al., 2024a), UniDepth (Piccinelli et al., 2024), ZeroDepth (Guizilini et al., 2023) and ZoeDepth (Bhat et al., 2023).”
• Continuous 3D Perception Model with Persistent State (2025, unknown): “Extensive research [7, 27, 42, 55, 58, 59, 85, 92] focuses on estimating monocular depth.”
• MoGe: Unlocking Accurate Monocular Geometry Estimation for Open-Domain Images with Optimal Training Supervision (2024, unknown): “UniDepth [38] proposed a self-promptable camera module that predicts dense camera representation to condition the subsequent depth estimation module.”
• Metric3D v2: A Versatile Monocular Geometric Foundation Model for Zero-Shot Metric Depth and Surface Normal Estimation (2024, unknown): “Con-currently, ZoeDepth [36], ZeroDepth [61], and UniDepth [62] apply varying strategies to tackle this challenge.”