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 2301.01352cs.LGcs.CV
收藏

WLD-Reg:一种数据相关的层内分集正则化器

WLD-Reg: A Data-dependent Within-layer Diversity Regularizer

Tampere University(坦佩雷大学);Aarhus University(奥胡斯大学)

Firas Laakom, Jenni Raitoharju, Alexandros Iosifidis, Moncef Gabbouj

摘要:

神经网络由以分层结构排列的多个层组成,所述分层结构利用基于梯度的优化联合训练,其中错误从最后一层反向传播回到第一层。在每个优化步骤中,给定层的神经元从属于层次结构的更高层的神经元接收反馈。在本文中,我们建议补充这种传统的“层间”的反馈与额外的“层内”的反馈,以鼓励同一层内的激活的多样性。为此,我们测量神经元输出之间的成对相似性,并使用它来建模层的整体多样性。我们提出了一个广泛的实证研究证实,所提出的方法提高了性能的几个国家的最先进的神经网络模型在多个任务。该代码可在\url{https://github.com/firasl/AAAI-23-WLD-Reg}上下载

英文摘要:

Neural networks are composed of multiple layers arranged in a hierarchical structure jointly trained with a gradient-based optimization, where the errors are back-propagated from the last layer back to the first one. At each optimization step, neurons at a given layer receive feedback from neurons belonging to higher layers of the hierarchy. In this paper, we propose to complement this traditional 'between-layer' feedback with additional 'within-layer' feedback to encourage the diversity of the activations within the same layer. To this end, we measure the pairwise similarity between the outputs of the neurons and use it to model the layer's overall diversity. We present an extensive empirical study confirming that the proposed approach enhances the performance of several state-of-the-art neural network models in multiple tasks. The code is publically available at \url{https://github.com/firasl/AAAI-23-WLD-Reg}

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。

↑