CORe50: a New Dataset and Benchmark for Continuous Object Recognition

TL;DR

The paper introduces CORe50, a new dataset and benchmark designed to evaluate continuous object recognition methods, addressing the lack of suitable datasets for lifelong learning in real-world applications.

Contribution

It provides a novel dataset and benchmark tailored for continuous object recognition, along with baseline approaches for various continuous learning scenarios.

Findings

CORe50 enables evaluation of lifelong learning algorithms.

Baseline methods demonstrate the benchmark's utility.

The dataset reflects real-world object recognition challenges.

Abstract

Continuous/Lifelong learning of high-dimensional data streams is a challenging research problem. In fact, fully retraining models each time new data become available is infeasible, due to computational and storage issues, while na\"ive incremental strategies have been shown to suffer from catastrophic forgetting. In the context of real-world object recognition applications (e.g., robotic vision), where continuous learning is crucial, very few datasets and benchmarks are available to evaluate and compare emerging techniques. In this work we propose a new dataset and benchmark CORe50, specifically designed for continuous object recognition, and introduce baseline approaches for different continuous learning scenarios.