Depth Anything 3: Recovering the Visual Space from Any Views

Haotong Lin; Sili Chen; Junhao Liew; Donny Y. Chen; Zhenyu Li; Guang Shi; Jiashi Feng; Bingyi Kang

返回文献探索

Paper · arXiv 2511.10647

Depth Anything 3: Recovering the Visual Space from Any Views

Haotong Lin, Sili Chen, Junhao Liew, Donny Y. Chen, Zhenyu Li, Guang Shi, Jiashi Feng, Bingyi Kang

103 upvotesNovember 13, 2025arXiv 预印本

AI 摘要

Depth Anything 3 (DA3) uses a plain transformer for geometry prediction from visual inputs, achieving state-of-the-art results in camera pose estimation, any-view geometry, visual rendering, and monocular depth estimation.

plain transformervanilla DINO encoderdepth-ray predictionteacher-student training paradigmvisual geometry benchmarkcamera pose estimationany-view geometryvisual renderingmonocular depth estimationstate-of-the-art (SOTA)VGGT

Abstract

We present Depth Anything 3 (DA3), a model that predicts spatially consistent geometry from an arbitrary number of visual inputs, with or without known camera poses. In pursuit of minimal modeling, DA3 yields two key insights: a single plain transformer (e.g., vanilla DINO encoder) is sufficient as a backbone without architectural specialization, and a singular depth-ray prediction target obviates the need for complex multi-task learning. Through our teacher-student training paradigm, the model achieves a level of detail and generalization on par with Depth Anything 2 (DA2). We establish a new visual geometry benchmark covering camera pose estimation, any-view geometry and visual rendering. On this benchmark, DA3 sets a new state-of-the-art across all tasks, surpassing prior SOTA VGGT by an average of 44.3% in camera pose accuracy and 25.1% in geometric accuracy. Moreover, it outperforms DA2 in monocular depth estimation. All models are trained exclusively on public academic datasets.