arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2108.09020cs.LGcs.CV
收藏

具有自然分布偏移的在线持续学习:基于可视化数据的实证研究

Online Continual Learning with Natural Distribution Shifts: An Empirical Study with Visual Data

Intel Labs(英特尔实验室)

Zhipeng Cai, Ozan Sener, Vladlen Koltun

摘要:

持续学习是在多个任务和环境中通过时间学习和保留知识的问题。研究主要集中在增量分类设置,其中新的任务/类以离散的时间间隔添加。这种“离线”设置不会评估代理有效和高效学习的能力,因为代理可以在添加任务时执行多个学习时期而没有任何时间限制。我们认为,“在线”持续学习,其中数据是一个单一的连续流没有任务边界,使评估信息保留和在线学习的功效。在在线持续学习中,每个传入的小批量数据首先用于测试,然后添加到训练集中,使问题真正在线。经过训练的模型随后会在历史数据上进行评估,以评估信息保留情况。我们介绍了一个新的基准在线持续视觉学习,表现出大规模和自然的分布偏移。通过大规模的分析,我们确定关键的和以前未观察到的现象,基于梯度的优化持续学习,并提出有效的策略,以改善基于梯度的在线持续学习与真实数据。源代码和数据集可在:https://github.com/IntelLabs/continuallearning。

英文摘要:

Continual learning is the problem of learning and retaining knowledge through time over multiple tasks and environments. Research has primarily focused on the incremental classification setting, where new tasks/classes are added at discrete time intervals. Such an "offline" setting does not evaluate the ability of agents to learn effectively and efficiently, since an agent can perform multiple learning epochs without any time limitation when a task is added. We argue that "online" continual learning, where data is a single continuous stream without task boundaries, enables evaluating both information retention and online learning efficacy. In online continual learning, each incoming small batch of data is first used for testing and then added to the training set, making the problem truly online. Trained models are later evaluated on historical data to assess information retention. We introduce a new benchmark for online continual visual learning that exhibits large scale and natural distribution shifts. Through a large-scale analysis, we identify critical and previously unobserved phenomena of gradient-based optimization in continual learning, and propose effective strategies for improving gradient-based online continual learning with real data. The source code and dataset are available in: https://github.com/IntelLabs/continuallearning.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑