arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2304.06309cs.CV
收藏

面向边缘设备的无需模型微调的分布外少样本学习

Out-of-distribution Few-shot Learning For Edge Devices without Model Fine-tuning

The Chinese University of Hong Kong(香港中文大学);Huawei Noah’s Ark Lab(华为诺亚方舟实验室)

Xinyun Zhang, Lanqing Hong

AI总结:

本文提出轻量即插即用的任务感知归一化模块TANO,无需反向传播即可实现边缘设备上分布外少样本任务的高效自适应,并在准确率、推理速度和模型大小上优于现有方法。

摘要:

通过利用有限数据定制深度学习网络的少样本学习(FSL)已成为在边缘设备上实现个性化用户体验的一种有前景的技术。然而,现有的FSL方法主要假设数据是独立同分布(IID)的,并且要么对每个任务使用计算反向传播更新,要么使用具有任务特定原型的通用模型。不幸的是,前一种方案对于缺乏设备端反向传播能力的边缘设备不可行,而后一种方案往往泛化能力有限,尤其是对于分布外(OOD)数据。本文提出了一种轻量级、即插即用的FSL模块,称为任务感知归一化(Task-aware Normalization,TANO),它能够在无需反向传播的情况下实现深度神经网络的高效且任务感知的自适应。TANO通过在元训练期间协调多组归一化统计量的更新来覆盖多个用户群体的特性,并自动为下游少样本任务识别合适的归一化组。因此,TANO提供了稳定但任务特定的归一化统计量估计,以缩小分布差距并实现高效的模型自适应。在域内和域外泛化实验中的结果表明,TANO在准确率、推理速度和模型大小方面均优于近期方法。此外,TANO在广泛使用的FSL基准和真实应用数据上取得了有前景的结果。

英文摘要:

Few-shot learning (FSL) via customization of a deep learning network with limited data has emerged as a promising technique to achieve personalized user experiences on edge devices. However, existing FSL methods primarily assume independent and identically distributed (IID) data and utilize either computational backpropagation updates for each task or a common model with task-specific prototypes. Unfortunately, the former solution is infeasible for edge devices that lack on-device backpropagation capabilities, while the latter often struggles with limited generalization ability, especially for out-of-distribution (OOD) data. This paper proposes a lightweight, plug-and-play FSL module called Task-aware Normalization (TANO) that enables efficient and task-aware adaptation of a deep neural network without backpropagation. TANO covers the properties of multiple user groups by coordinating the updates of several groups of the normalization statistics during meta-training and automatically identifies the appropriate normalization group for a downstream few-shot task. Consequently, TANO provides stable but task-specific estimations of the normalization statistics to close the distribution gaps and achieve efficient model adaptation. Results on both intra-domain and out-of-domain generalization experiments demonstrate that TANO outperforms recent methods in terms of accuracy, inference speed, and model size. Moreover, TANO achieves promising results on widely-used FSL benchmarks and data from real applications.

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。

↑