Online Open World Recognition

TL;DR

This paper advances open world recognition by integrating online metric learning, local learning, and dynamic threshold estimation, enabling recognition systems to adapt continuously to new classes and data in large-scale, evolving environments.

Contribution

It introduces a comprehensive online learning framework that incorporates metric learning, confidence threshold adaptation, and local learning for open world recognition.

Findings

Outperforms non-online methods on large-scale datasets

Effectively detects and incorporates new classes over time

Enhances recognition accuracy with dynamic, incremental updates

Abstract

As we enter into the big data age and an avalanche of images have become readily available, recognition systems face the need to move from close, lab settings where the number of classes and training data are fixed, to dynamic scenarios where the number of categories to be recognized grows continuously over time, as well as new data providing useful information to update the system. Recent attempts, like the open world recognition framework, tried to inject dynamics into the system by detecting new unknown classes and adding them incrementally, while at the same time continuously updating the models for the known classes. incrementally adding new classes and detecting instances from unknown classes, while at the same time continuously updating the models for the known classes. In this paper we argue that to properly capture the intrinsic dynamic of open world recognition, it is necessary to add to these aspects (a) the incremental learning of the underlying metric, (b) the incremental estimate of confidence thresholds for the unknown classes, and (c) the use of local learning to precisely describe the space of classes. We extend three existing metric learning algorithms towards these goals by using online metric learning. Experimentally we validate our approach on two large-scale datasets in different learning scenarios. For all these scenarios our proposed methods outperform their non-online counterparts. We conclude that local and online learning is important to capture the full dynamics of open world recognition.