Register now Registered users can see all details and apply online for the opportunities.
Self-Supervised Learning for Robotic Navigation
Learning representations for navigation in a self-supervised manner.
Keywords: Self-supervised, deep learning, navigation, reinforcement learning
While there are massive labeled data sets available for computer vision researchers, they cannot directly be applied to robotics. However, robots can interact with their environment and learn about the consequences of their actions. Your task is to formalize this abstract notion such that the robot can learn representations for navigation in a self-supervised fashion. Depending on your progress you will use the learned representation to facilitate reinforcement learning of navigation tasks.
While there are massive labeled data sets available for computer vision researchers, they cannot directly be applied to robotics. However, robots can interact with their environment and learn about the consequences of their actions. Your task is to formalize this abstract notion such that the robot can learn representations for navigation in a self-supervised fashion. Depending on your progress you will use the learned representation to facilitate reinforcement learning of navigation tasks.
Not specified
Please send your CV and transcript to:
Mathias Gehrig, mgehrig (at) ifi (dot) uzh (dot) ch