arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2103.16889cs.CV
收藏

神经传递和结构自适应联合学习的图像识别

Joint Learning of Neural Transfer and Architecture Adaptation for Image Recognition

Sun Yat-sen University(中山大学);DarkMatter AI Research(DarkMatter AI 研究院)

Guangrun Wang, Liang Lin, Rongcong Chen, Guangcong Wang, Jiqi Zhang

摘要:

当前最先进的视觉识别系统通常依赖于以下流水线:(a)在大规模数据集上预训练神经网络(例如,ImageNet)和(b)在较小的特定任务数据集上微调网络权重。这样的流水线假设唯一的权重自适应能够基于固定架构适合于所有域的强假设将网络能力从一个域转移到另一个域。然而,具有不同识别目标的每个域可能需要不同级别/路径的特征层次结构,其中一些神经元可能变得冗余,并且一些其他神经元被重新激活以形成新的网络结构。在这项工作中,我们证明了与现有的图像识别流水线相比,动态适应为每个域任务定制的网络架构以及权重微调在效率和有效性方面都有好处,而现有的图像识别流水线只调整权重,而不管架构如何。我们的方法可以很容易地推广到一个无监督的范例,通过在源域任务中用自监督学习代替超网训练,并在下游任务中执行线性评估。这进一步提高了我们的方法的搜索效率。此外,我们还提供了原则和实证分析,以解释为什么我们的方法通过调查现有的神经结构搜索的无效性。我们发现,保持网络结构和权重的联合分布是很重要的。这种分析不仅有利于图像识别,还为制作神经网络提供了见解。在人物再识别、年龄估计、性别识别、图像分类和无监督领域自适应等五个典型图像识别任务上的实验证明了该方法的有效性。

英文摘要:

Current state-of-the-art visual recognition systems usually rely on the following pipeline: (a) pretraining a neural network on a large-scale dataset (e.g., ImageNet) and (b) finetuning the network weights on a smaller, task-specific dataset. Such a pipeline assumes the sole weight adaptation is able to transfer the network capability from one domain to another domain, based on a strong assumption that a fixed architecture is appropriate for all domains. However, each domain with a distinct recognition target may need different levels/paths of feature hierarchy, where some neurons may become redundant, and some others are re-activated to form new network structures. In this work, we prove that dynamically adapting network architectures tailored for each domain task along with weight finetuning benefits in both efficiency and effectiveness, compared to the existing image recognition pipeline that only tunes the weights regardless of the architecture. Our method can be easily generalized to an unsupervised paradigm by replacing supernet training with self-supervised learning in the source domain tasks and performing linear evaluation in the downstream tasks. This further improves the search efficiency of our method. Moreover, we also provide principled and empirical analysis to explain why our approach works by investigating the ineffectiveness of existing neural architecture search. We find that preserving the joint distribution of the network architecture and weights is of importance. This analysis not only benefits image recognition but also provides insights for crafting neural networks. Experiments on five representative image recognition tasks such as person re-identification, age estimation, gender recognition, image classification, and unsupervised domain adaptation demonstrate the effectiveness of our method.

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。

↑