引导动态感知的 3D 视觉表征用于可扩展机器人学习
Bootstrap Dynamic-Aware 3D Visual Representation for Scalable Robot Learning
把状态预测建模为扩散生成过程、在共享潜空间联合学习前向与逆向动力学，为视觉表征注入可操作的因果转移结构；特征差分与逆一致性监督可防止动作学习中的特征泄漏，这一思路可直接迁移到任何需要动作条件表征的视觉预训练中。
原文依据 AFRO casts state prediction as a generative diffusion process and jointly models forward and inverse dynamics in a shared latent space
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论尽管在识别与分割任务上表现强劲，当前的三维视觉预训练方法在机器人操作任务上往往表现不佳。我们将这一差距归因于两个因素：缺乏状态-动作-状态动力学建模，以及显式几何重建带来的不必要冗余。我们提出 AFRO，一个自监督框架，无需动作或重建监督即可学习动态感知的三维表征。AFRO 将状态预测转化为生成式扩散过程，并在共享潜空间中联合建模前向与逆向动力学，以捕捉因果转移结构。为防止动作学习中的特征泄漏，我们采用特征差分与逆一致性监督，提升视觉特征的质量与稳定性。与 Diffusion Policy 结合后，AFRO 在 16 个仿真任务和 4 个真实世界任务上大幅提升操作成功率，优于现有预训练方法。该框架还能随数据规模和任务复杂度良好扩展。定性可视化表明，AFRO 学到了语义丰富且具有判别性的特征，为机器人领域的三维表征学习提供了一种有效的预训练方案。项目主页：https://kolakivy.github.io/AFRO/
英文原文摘要 1270 字符
Despite strong results on recognition and segmentation, current 3D visual pre-training methods often underperform on robotic manipulation. We attribute this gap to two factors: the lack of state-action-state dynamics modeling and the unnecessary redundancy of explicit geometric reconstruction. We introduce AFRO, a self-supervised framework that learns dynamics-aware 3D representations without action or reconstruction supervision. AFRO casts state prediction as a generative diffusion process and jointly models forward and inverse dynamics in a shared latent space to capture causal transition structure. To prevent feature leakage in action learning, we employ feature differencing and inverse-consistency supervision, improving the quality and stability of visual features. When combined with Diffusion Policy, AFRO substantially increases manipulation success rates across 16 simulated and 4 real-world tasks, outperforming existing pre-training approaches. The framework also scales favorably with data volume and task complexity. Qualitative visualizations indicate that AFRO learns semantically rich, discriminative features, offering an effective pre-training solution for 3D representation learning in robotics. Project page: https://kolakivy.github.io/AFRO/