LA-Pose：潜变量动作预训练遇上姿态估计
LA-Pose: Latent Action Pretraining Meets Pose Estimation
将逆动力学自监督预训练得到的潜变量动作特征作为相机姿态估计器的输入，仅需少量高质量3D标注微调即可实现高精度、可泛化的前馈姿态预测，为视觉里程计和自动驾驶感知提供了一种可扩展的预训练范式。
原文依据 This paper revisits camera pose estimation through the lens of self-supervised pretraining, focusing on inverse-dynamics pretraining as a scalable alternative to the current trend of fully supervised training with 3D annotations.
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论本文从自监督预训练的视角重新审视相机姿态估计，聚焦于逆动力学预训练，将其作为当前完全监督训练与3D标注趋势的可扩展替代方案。具体而言，我们采用逆向和正向动力学模型来学习潜变量动作表示，类似于在大规模驾驶视频上训练的Genie。我们的想法简单而有效。现有方法使用潜变量动作的原始能力，即作为世界模型的动作条件或策略网络中机器人动作参数的代理。我们的方法称为LA-Pose，将潜变量动作特征重新用作相机姿态估计器的输入，并在有限的高质量3D标注上微调。这种公式化方法能够实现准确且可泛化的姿态预测，同时保持前馈效率。在驾驶基准上的大量实验表明，LA-Pose在使用少几个数量级的标注数据的情况下，达到了与最优方法相当甚至更优的性能。具体而言，在Waymo和PandaSet基准上，LA-Pose的姿态精度比最近的前馈方法高出10%以上。据我们所知，这项工作首次展示了逆动力学自监督学习在姿态估计中的威力。
英文原文摘要 1310 字符
This paper revisits camera pose estimation through the lens of self-supervised pretraining, focusing on inverse-dynamics pretraining as a scalable alternative to the current trend of fully supervised training with 3D annotations. Concretely, we employ inverse- and forward-dynamics models to learn latent action representations, similar to Genie from large-scale driving videos.Our idea is simple yet effective. Existing methods use latent actions in their original capacity, that is, as action conditioning of world-models or as proxies of robot action parameters in policy networks.Our method, dubbed LA-Pose, repurposes the latent action features as inputs to a camera pose estimator, finetuned on a limited set of high-quality 3D annotations.This formulation enables accurate and generalizable pose prediction while maintaining feed-forward efficiency. Extensive experiments on driving benchmarks show that LA-Pose achieves competitive and even superior performance to state-of-the-art methods while using orders of magnitude less labeled data. Concretely, on the Waymo and PandaSet benchmarks, LA-Pose achieves over 10% higher pose accuracy than recent feed-forward methods.To our knowledge, this work is the first to demonstrate the power of inverse-dynamics self-supervised learning for pose estimation.