Depth Anything: Unleashing the Power of Large-Scale Unlabeled Data

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Yang, Lihe, Kang, Bingyi, Huang, Zilong, Xu, Xiaogang, Feng, Jiashi, Zhao, Hengshuang
Formato: Preprint
Publicado: 2024
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866913304280039424
author Yang, Lihe
Kang, Bingyi
Huang, Zilong
Xu, Xiaogang
Feng, Jiashi
Zhao, Hengshuang
author_facet Yang, Lihe
Kang, Bingyi
Huang, Zilong
Xu, Xiaogang
Feng, Jiashi
Zhao, Hengshuang
contents This work presents Depth Anything, a highly practical solution for robust monocular depth estimation. Without pursuing novel technical modules, we aim to build a simple yet powerful foundation model dealing with any images under any circumstances. To this end, we scale up the dataset by designing a data engine to collect and automatically annotate large-scale unlabeled data (~62M), which significantly enlarges the data coverage and thus is able to reduce the generalization error. We investigate two simple yet effective strategies that make data scaling-up promising. First, a more challenging optimization target is created by leveraging data augmentation tools. It compels the model to actively seek extra visual knowledge and acquire robust representations. Second, an auxiliary supervision is developed to enforce the model to inherit rich semantic priors from pre-trained encoders. We evaluate its zero-shot capabilities extensively, including six public datasets and randomly captured photos. It demonstrates impressive generalization ability. Further, through fine-tuning it with metric depth information from NYUv2 and KITTI, new SOTAs are set. Our better depth model also results in a better depth-conditioned ControlNet. Our models are released at https://github.com/LiheYoung/Depth-Anything.
format Preprint
id arxiv_https___arxiv_org_abs_2401_10891
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Depth Anything: Unleashing the Power of Large-Scale Unlabeled Data
Yang, Lihe
Kang, Bingyi
Huang, Zilong
Xu, Xiaogang
Feng, Jiashi
Zhao, Hengshuang
Computer Vision and Pattern Recognition
This work presents Depth Anything, a highly practical solution for robust monocular depth estimation. Without pursuing novel technical modules, we aim to build a simple yet powerful foundation model dealing with any images under any circumstances. To this end, we scale up the dataset by designing a data engine to collect and automatically annotate large-scale unlabeled data (~62M), which significantly enlarges the data coverage and thus is able to reduce the generalization error. We investigate two simple yet effective strategies that make data scaling-up promising. First, a more challenging optimization target is created by leveraging data augmentation tools. It compels the model to actively seek extra visual knowledge and acquire robust representations. Second, an auxiliary supervision is developed to enforce the model to inherit rich semantic priors from pre-trained encoders. We evaluate its zero-shot capabilities extensively, including six public datasets and randomly captured photos. It demonstrates impressive generalization ability. Further, through fine-tuning it with metric depth information from NYUv2 and KITTI, new SOTAs are set. Our better depth model also results in a better depth-conditioned ControlNet. Our models are released at https://github.com/LiheYoung/Depth-Anything.
title Depth Anything: Unleashing the Power of Large-Scale Unlabeled Data
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2401.10891