Self-supervised learning for autonomous vehicles perception: A conciliation between analytical and learning methods

TL;DR

This paper reviews self-supervised learning techniques for autonomous vehicle perception, highlighting their advantages over supervised methods in reducing manual labeling and adapting to changing environments.

Contribution

It provides a formal overview of SSL concepts for autonomous vehicles and offers guidelines for developing new SSL frameworks and addressing future challenges.

Findings

SSL techniques can replace manual labeling in autonomous perception tasks

SSL enables learning in varying environments and over time

The paper identifies key challenges in designing SSL systems for autonomous vehicles

Abstract

Nowadays, supervised deep learning techniques yield the best state-of-the-art prediction performances for a wide variety of computer vision tasks. However, such supervised techniques generally require a large amount of manually labeled training data. In the context of autonomous vehicles perception, this requirement is critical, as the distribution of sensor data can continuously change and include several unexpected variations. It turns out that a category of learning techniques, referred to as self-supervised learning (SSL), consists of replacing the manual labeling effort by an automatic labeling process. Thanks to their ability to learn on the application time and in varying environments, state-of-the-art SSL techniques provide a valid alternative to supervised learning for a variety of different tasks, including long-range traversable area segmentation, moving obstacle instance segmentation, long-term moving obstacle tracking, or depth map prediction. In this tutorial-style article, we present an overview and a general formalization of the concept of self-supervised learning (SSL) for autonomous vehicles perception. This formalization provides helpful guidelines for developing novel frameworks based on generic SSL principles. Moreover, it enables to point out significant challenges in the design of future SSL systems.