雅虎香港 搜尋

搜尋結果

    • Neural simulator

      • Discover Wayve Ghost Gym, a neural simulator that creates photorealistic 4D worlds and thousands of simulated scenarios for training, testing, and debugging our end-to-end AI driving models thoroughly in an offline, virtual environment.
  1. 其他人也問了

  2. 2023年12月21日 · Ghost Gym is a closed-loop neural simulator built with data-driven techniques to model the environment and the actors within it. We achieve this by using learnt neural rendering techniques that model the geometry, motion, and appearance of the scene, as well as how our robot would interact with the given environment.

  3. 2024年5月7日 · Ghost Gym: A Neural Simulator for Autonomous Driving Discover Wayve Ghost Gym, a neural simulator that creates photorealistic 4D worlds and thousands of simulated scenarios for training, testing, and debugging our end-to-end AI driving models thoroughly in an offline, virtual environment.

    • 概览
    • 参考

    本文旨在解读,学习Wayve的神经网络仿真器工具GhostGym,原文已联接[1]。GhostGym的核心是基于神经网络来模拟无人车与数字世界的整个闭环仿真过程。

    仿真是无人车领域实际路测之外的重要组成部分。Wayve使用先进的神经网络来渲染,创建逼真的数字世界,整个仿真世界就是自车驾驶模型之外的另外一个大模型。笔者这里猜测神经网络应该是基于GAIA-1[2]的世界模型构建的。可见Wayve这家公司真的很ML driven。Wayve也提到,他们公司已经开始探索利用神经网络来代替传统的基于rule-based的仿真框架。

    Ghost Gym是网络驱动基于闭环框架的模拟器,模型负责建模运动轨迹和感知场景,以及每个物体间的交互。开环仿真不会考虑无人车仿真出的新轨迹和数字世界的交互,而Ghost Gym的闭环仿真则基于模型仿真出交互结果。因为Wayve的新版自动驾驶系统是基于端到端的大模型,所以Ghost Gym同样也要和这个大模型兼容,给出高质量的仿真结果来作为自动驾驶模型的输入。

    Ghost Gym的组成模块包括:1)高还原度的自车车机系统。2)精确的汽车动力学模型。3) 高质量的神经网络来渲染输出新的仿真感知场景。但是笔者认为前两个模块是任何仿真系统都应该普遍具备的。而且Wayve对于这两个模块的构建没有特殊性(应该没用模型)。

  4. BOO! 👻 Introducing Wayve Ghost Gym, a closed-loop neural simulator where we can train, test and debug our end-to-end AI driving models. Come and take a look inside 🫴 https://t.co/8NmqQeCwnQ #GhostGym #EmbodiedAI #AutonomousVehicle #Selfdriving pic.twitter.com/feJUQgPd2e. — Wayve (@wayve_ai) December 21, 2023. twitter.com. 1. Sort by: Add a Comment.

  5. 2024年4月17日 · Tested in Wayve’s neural simulator, Ghost Gym, and on public roads, LINGO-2 gives a strong initial signal of what can be achieved by aligning linguistic explanations and decision-making.

  6. 2022年9月26日 · Wayve, a driverless-car startup based in London, has made a machine-learning model that can drive two different types of vehicle: a passenger car and a delivery van. It is the first time the...