ETH Zurich
A leading Swiss research university known for its strengths in science and engineering, with significant research contributions in areas including AI safety and machine learning.
Timeline (last 90 days)
Daisy Dent is a forest ecologist and senior scientist at ETH Zurich.
The study found that both attacks against agents and the defenses built for them are extremely challenging, and it is difficult for defenders to completely plug vulnerabilities without sacrificing functionality.
ETH Zurich conducted systematic testing of tool-using agents, covering 97 tasks close to real-world scenarios and designing 629 safety tests.
In ETH Zurich's ANYmal series of work, terrain difficulty, velocity targets, and other parameters are automatically upgraded as training progresses (applying curriculum learning).
Landmark works such as OpenAI's training of a robotic hand to solve a Rubik's Cube and ETH Zurich's training of quadruped robots both use domain randomization as a core technique.
RSL-RL is a lightweight reinforcement learning framework developed by ETH Zurich's Robotic Systems Lab specifically for legged robots, and is deeply integrated with Isaac Gym.
ETH Zurich's ANYmal quadruped robot team used the Sim2Real method to enable the robot to learn adaptive walking on rough terrain, with the policy trained entirely in simulation.
ETH Zurich's ANYmal quadruped robot and multiple studies from UC Berkeley have demonstrated that carefully designed simulation training strategies can be deployed to physical robots with almost zero adjustments.
OpenAI's use of the sim-to-real method to train a robotic hand to solve a Rubik's Cube and ETH Zurich's ANYmal quadruped robot walking on outdoor terrain are both landmark achievements of sim-to-real.
2022年以来，ETH Zurich的ANYmal团队和MIT的Mini Cheetah团队展示了在崎岖地形上稳定行走的四足机器人，策略均在仿真中并行训练后迁移到真实硬件
2 more timeline events
All Facts (12)
ETH Zurich's ANYmal quadruped robot team used the Sim2Real method to enable the robot to learn adaptive walking on rough terrain, with the policy trained entirely in simulation.
60%UnverifiedDaisy Dent is a forest ecologist and senior scientist at ETH Zurich.
50%UnverifiedThe study found that both attacks against agents and the defenses built for them are extremely challenging, and it is difficult for defenders to completely plug vulnerabilities without sacrificing functionality.
50%UnverifiedETH Zurich conducted systematic testing of tool-using agents, covering 97 tasks close to real-world scenarios and designing 629 safety tests.
50%UnverifiedIn ETH Zurich's ANYmal series of work, terrain difficulty, velocity targets, and other parameters are automatically upgraded as training progresses (applying curriculum learning).
50%UnverifiedLandmark works such as OpenAI's training of a robotic hand to solve a Rubik's Cube and ETH Zurich's training of quadruped robots both use domain randomization as a core technique.
50%UnverifiedRSL-RL is a lightweight reinforcement learning framework developed by ETH Zurich's Robotic Systems Lab specifically for legged robots, and is deeply integrated with Isaac Gym.
50%UnverifiedETH Zurich's ANYmal quadruped robot and multiple studies from UC Berkeley have demonstrated that carefully designed simulation training strategies can be deployed to physical robots with almost zero adjustments.
50%UnverifiedOpenAI's use of the sim-to-real method to train a robotic hand to solve a Rubik's Cube and ETH Zurich's ANYmal quadruped robot walking on outdoor terrain are both landmark achievements of sim-to-real.
50%Unverified2022年以来，ETH Zurich的ANYmal团队和MIT的Mini Cheetah团队展示了在崎岖地形上稳定行走的四足机器人，策略均在仿真中并行训练后迁移到真实硬件
50%UnverifiedETH Zurich 的 Agile Autonomy 项目通过特权学习框架实现了无人机在复杂环境中的高速飞行
50%UnverifiedETH Zurich 的 Agile Autonomy 项目通过特权学习（Privileged Learning）框架实现了无人机在复杂环境中的高速飞行
50%