Visual Pre-training for Navigation: What Can We Learn from Noise?

Yanwei Wang     Ching-Yun Ko     Pulkit Agrawal

Computer Science and Artificial Intelligence Laboratory (CSAIL)
Massachusetts Institute of Technology

In visual navigation, one powerful paradigm is to predict actions from observations directly. Training such an end-to-end system allows representations that are useful for downstream tasks to emerge automatically. However, the lack of inductive bias makes this system data-hungry. We hypothesize a sufficient representation of the current view and the goal view for a navigation policy can be learned by predicting the location and size of a crop of the current view that corresponds to the goal. We further show that training such random crop prediction in a self-supervised fashion purely on synthetic noise images transfers well to natural home images. The learned representation can then be bootstrapped to learn a navigation policy efficiently with little interaction data.