arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2409.02486cs.CVcs.AI

通过元初始化提升零样本跨数据集单图像室内深度估计的泛化能力

Boosting Generalizability towards Zero-Shot Cross-Dataset Single-Image Indoor Depth by Meta-Initialization

Institution1（机构1）；Institution2（机构2）

Cho-Ying Wu, Yiqi Zhong, Junying Wang, Ulrich Neumann

备注：
IROS 2024. The version supersedes 2305.07269. arXiv admin note: text overlap with arXiv:2305.07269
阅读 PDF
arXiv 页面 ↗HTML 原文 ↗

AI总结：

本文针对室内单图像深度估计的跨数据集泛化问题，提出将每个RGB-D小批量视为细粒度任务的元学习方法，通过元初始化获得更好先验并提升零样本跨数据集性能。

摘要：

室内机器人依赖深度信息执行导航或障碍物检测等任务，单图像深度估计被广泛用于辅助感知。大多数室内单图像深度预测方法较少关注模型对未见数据集的泛化能力，而这对系统部署中的真实环境鲁棒性至关重要。本工作利用基于梯度的元学习来获得更高的零样本跨数据集推理泛化能力。与已有大量研究的、与显式类别标签相关联的图像分类元学习不同，连续深度值与高度变化的室内环境（涉及物体布局和场景构成）之间不存在明确的任务边界。我们提出细粒度任务，在元学习框架中将每个 RGB-D 小批量视为一个任务。我们首先表明，在有限数据上我们的方法能够诱导出更好的先验（RMSE 最高改善 27.8%）。随后，在元学习初始化基础上进行微调，始终优于未采用元方法的基线。以泛化为目标，我们提出了零样本跨数据集协议，并验证了由元初始化带来的更高泛化能力，该方法可作为许多现有深度估计方法的一个简单且有用的插件。这项处于深度估计与元学习交叉领域的工作有望推动两个研究方向更接近实际机器人和机器感知应用。

英文摘要：

Indoor robots rely on depth to perform tasks like navigation or obstacle detection, and single-image depth estimation is widely used to assist perception. Most indoor single-image depth prediction focuses less on model generalizability to unseen datasets, concerned with in-the-wild robustness for system deployment. This work leverages gradient-based meta-learning to gain higher generalizability on zero-shot cross-dataset inference. Unlike the most-studied meta-learning of image classification associated with explicit class labels, no explicit task boundaries exist for continuous depth values tied to highly varying indoor environments regarding object arrangement and scene composition. We propose fine-grained task that treats each RGB-D mini-batch as a task in our meta-learning formulation. We first show that our method on limited data induces a much better prior (max 27.8% in RMSE). Then, finetuning on meta-learned initialization consistently outperforms baselines without the meta approach. Aiming at generalization, we propose zero-shot cross-dataset protocols and validate higher generalizability induced by our meta-initialization, as a simple and useful plugin to many existing depth estimation methods. The work at the intersection of depth and meta-learning potentially drives both research to step closer to practical robotic and machine perception usage.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。

↑