统一感知与动作：一种面向机器人动作生成的隐式视觉思维链混合模态流水线
Unifying Perception and Action: A Hybrid-Modality Pipeline with Implicit Visual Chain-of-Thought for Robotic Action Generation
VITA 将视觉预测与动作生成统一到共享离散潜空间，用自回归 token 同时解码未来帧和动作，为视觉研究者提供了一种把视觉动态作为归纳偏置注入策略学习、缓解模态鸿沟与训练不稳定的可复用范式。
原文依据 learns a shared discrete latent space for vision and action, enabling joint modeling of perception and motor control
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论基于思维链（CoT）的视觉-语言-动作（VLA）模型凭借其显著的感知理解能力，在推进通用机器人智能体方面取得了显著成功。近期，由于纯文本思维链难以在复杂空间环境中充分捕捉场景细节，一种极具前景的策略是利用视觉先验来引导机器人动作生成。然而，这些策略面临两个固有挑战：（i）视觉观测与底层动作之间的模态鸿沟；（ii）视觉预测与动作生成之间的目标竞争导致训练不稳定。为解决这些挑战，我们提出了视觉集成轨迹对齐（VITA）框架，该框架学习视觉与动作共享的离散潜空间，从而实现感知与运动控制的联合建模。VITA 引入了一种隐式视觉思维链：自回归生成的 token 被同时解码为未来帧预测和机器人动作，从而将视觉动态内化为运动规划的归纳偏置。在仿真和真实环境中的大量实验表明，VITA 达到了最优性能。在 CALVIN、LIBERO 和 SimplerEnv 上，VITA 分别比现有基线提升了 14.5%、9.6% 和 12.1%。此外，VITA 在六个真实世界任务中平均成功率达到 80.5%，展示了其作为通用机器人操作模型的潜力。
英文原文摘要 1449 字符
Vision-Language-Action (VLA) models built upon Chain-of-Thought (CoT) have achieved remarkable success in advancing general-purpose robotic agents, owing to its significant perceptual comprehension. Recently, since text-only CoT struggles to adequately capture scene details in complex spatial environments, a highly promising strategy involves leveraging visual priors to guide robotic action generation. Nevertheless, these strategies face two inherent challenges: (i) a modality gap between visual observations and low-level actions, and (ii) unstable training due to competing objectives between visual prediction and action generation. To address these challenges, we propose a Vision-Integrated Trajectory Alignment (VITA) framework that learns a shared discrete latent space for vision and action, enabling joint modeling of perception and motor control. VITA introduces a implicit visual CoT: autoregressively generated tokens is simultaneously decoded into future frames predictions and robot actions, thereby internalizing visual dynamics as an inductive bias for motion planning. Extensive experiments on simulated and real-world environments demonstrate state-of-the-art performance. VITA improves 14.5%, 9.6% and 12.1% over existing baselines on CALVIN, LIBERO and SimplerEnv. Furthermore, VITA attains an average success rate of 80.5% across six real-world tasks, demonstrating its potential as a generalist robotic manipulation model.