Sergey Levine and colleagues, 2016
We develop a method that can be used to learn policies that map raw image observations directly to torques at the robot’s motors.
Don't Just Read it – Explore It
We develop a method that can be used to learn policies that map raw image observations directly to torques at the robot’s motors.
We propose a novel neural architecture Transformer-XL that enables learning dependency beyond a fixed length without disrupting temporal coherence.
We adapt the ideas underlying the success of Deep Q-Learning to the continuous action domain.
Dealing with sparse rewards is one of the biggest challenges in Reinforcement Learning.
We describe a simple approach for this task based on a transformer.
The actor aims to maximize expected reward while also maximizing entropy.
We present high quality image synthesis results using diffusion probabilistic models.
We apply them in the latent space of powerful pretrained autoencoders.
We propose an alternative generator architecture for generative adversarial networks, borrowing from style transfer literature.
Learning useful representations without supervision remains a key challenge in machine learning.
We investigate conditional adversarial networks as a general-purpose solution to image-to-image translation problems.
A pure transformer applied directly to sequences of image patches can perform very well on image classification tasks.