CLAD: A realistic Continual Learning benchmark for Autonomous Driving

TL;DR

This paper introduces CLAD, a new realistic continual learning benchmark for autonomous driving using SODA10M, focusing on object classification and detection challenges, and analyzes current methods and future directions.

Contribution

The paper presents CLAD, a novel benchmark for continual learning in autonomous driving, and provides a comprehensive analysis of existing benchmarks and techniques.

Findings

CLAD includes two benchmarks: CLAD-C for classification and CLAD-D for detection.

Top methods in CLAD challenge face class and domain incremental difficulties.

Survey of techniques reveals current challenges and promising future research directions.

Abstract

In this paper we describe the design and the ideas motivating a new Continual Learning benchmark for Autonomous Driving (CLAD), that focuses on the problems of object classification and object detection. The benchmark utilises SODA10M, a recently released large-scale dataset that concerns autonomous driving related problems. First, we review and discuss existing continual learning benchmarks, how they are related, and show that most are extreme cases of continual learning. To this end, we survey the benchmarks used in continual learning papers at three highly ranked computer vision conferences. Next, we introduce CLAD-C, an online classification benchmark realised through a chronological data stream that poses both class and domain incremental challenges; and CLAD-D, a domain incremental continual object detection benchmark. We examine the inherent difficulties and challenges posed by the benchmark, through a survey of the techniques and methods used by the top-3 participants in a CLAD-challenge workshop at ICCV 2021. We conclude with possible pathways to improve the current continual learning state of the art, and which directions we deem promising for future research.