nirviaje/deeplearningrobotics

for integrating Deep Learning and Robotics

deeplearningrobotics's Issues

DQN-Sandbox

https://github.com/A-Noctua/DQN-Sandbox

Deep Q Learning application in robotics
http://rll.berkeley.edu/deeplearningrobotics/index.html

看来作者已经弃坑了

哈哈蛤

the bitdeli

rt,.

Meta-World

Meta-World is an open-source simulated benchmark for meta-reinforcement learning and multi-task learning consisting of 50 distinct robotic manipulation tasks.

• https://github.com/rlworkgroup/metaworld
• https://arxiv.org/pdf/1910.10897.pdf

The actions in this space

range between −1 and 1.

For all tasks, the robot must either manipulate

• one object with a variable goal position, or
• manipulate two objects with a fixed goal position.

The observation space is represented as

• a 6-tuple of the 3D Cartesian positions of the end-effector,
• a normalized measurement of how open the gripper is,
• object
  • the 3D position of the first object,
  • the quaternion of the first object,
  • the 3D position of the second object,
  • the quaternion of the second object,
• all of the previous measurements in the environment, and
  • finally the 3D position of the goal.

If there is no second object or the goal is not meant to be included in the observation, then the quantities corresponding to them are zeroed out. The observation space is always 39 dimensional.

Building block for intelligence

• Building block for intelligence
  • Briefly to information theory
  • Universal Approximation Theorem
  • manifold hypothesis[Mitter, 2016] (eg. random embedding), etc.
  • inductive bias, No-Free-Lunch
  • Brain intelligence
  • MLP/CNN/LSTM/Attention and more
  • Around NN: data, arch (w optimizer), and representation
  • Paradigm for learning system