Register now After registration you will be able to apply for this opportunity online.
This opportunity is not published. No applications will be accepted.
Visual Representation Learning for Efficient Deep Reinforcement Learning
Visual Representation Learning for Efficient Deep Reinforcement Learning
The study of end-to-end deep learning in computer vision has mainly focused on developing useful object representations for image classification, object detection, or semantic segmentation. Recent work has shown that it is possible to learn temporally and geometrically aligned keypoints given only videos, and the object keypoints learned via unsupervised learning manners can be useful for efficient control and reinforcement learning.
The study of end-to-end deep learning in computer vision has mainly focused on developing useful object representations for image classification, object detection, or semantic segmentation. Recent work has shown that it is possible to learn temporally and geometrically aligned keypoints given only videos, and the object keypoints learned via unsupervised learning manners can be useful for efficient control and reinforcement learning.
The goal of this project is to find out if it is possible to learn useful features or intermediate representations for controlling mobile robots at high speed. For example, can we use the Transporter (a neural network architecture) to find useful features in an autonomous car racing environment? if so, can we use these features to discover an optimal control policy via deep reinforcement learning?
The goal of this project is to find out if it is possible to learn useful features or intermediate representations for controlling mobile robots at high speed. For example, can we use the Transporter (a neural network architecture) to find useful features in an autonomous car racing environment? if so, can we use these features to discover an optimal control policy via deep reinforcement learning?
Yunlong Song (song@ifi.uzh.ch), Jiaxu Xing (jixing@ifi.uzh.ch)
Yunlong Song (song@ifi.uzh.ch), Jiaxu Xing (jixing@ifi.uzh.ch)