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 1907.00542cs.LGeess.ASeess.IVstat.ML
收藏

基于余弦相似性的对抗过程

Cosine similarity-based adversarial process

University of Seoul(首尔大学)

Hee-Soo Heo, Jee-weon Jung, Hye-jin Shim, IL-Ho Yang, Ha-Jin Yu

摘要:

两个深度神经网络之间的对抗过程是训练鲁棒模型的一种很有前途的方法。在本文中,我们提出了一个使用余弦相似性的对抗过程,而传统的对抗过程是基于反向分类交叉熵(CCE)。当用于训练识别模型时,对抗过程引起两个辨别模型的竞争;一个用于主要任务,如说话人辨识或图像识别,另一个用于辅助任务,如通道识别或域识别。特别是,对抗过程通过消除输入中的辅助信息来降低辅助模型的性能,假设输入中的辅助信息可能会降低主模型的性能。传统的对抗过程最大化子模型的CCE,从而降低性能。我们已经研究了一个框架,通过应用这样一个对抗过程消除通道或域信息(辅助信息)来训练鲁棒的判别模型。然而,我们通过实验发现,使用最大化CCE的过程并不能保证子模型的性能下降。相反,在所提出的使用余弦相似性的对抗过程中,通过搜索与辅助模型正交的特征空间,可以更有效地降低辅助模型的性能。通过对说话人辨识和图像识别的实验表明,本文提出的方法能够有效地提高辅助模型的输出与输入的独立性,从而提高了主模型的识别性能。

英文摘要:

An adversarial process between two deep neural networks is a promising approach to train a robust model. In this paper, we propose an adversarial process using cosine similarity, whereas conventional adversarial processes are based on inverted categorical cross entropy (CCE). When used for training an identification model, the adversarial process induces the competition of two discriminative models; one for a primary task such as speaker identification or image recognition, the other one for a subsidiary task such as channel identification or domain identification. In particular, the adversarial process degrades the performance of the subsidiary model by eliminating the subsidiary information in the input which, in assumption, may degrade the performance of the primary model. The conventional adversarial processes maximize the CCE of the subsidiary model to degrade the performance. We have studied a framework for training robust discriminative models by eliminating channel or domain information (subsidiary information) by applying such an adversarial process. However, we found through experiments that using the process of maximizing the CCE does not guarantee the performance degradation of the subsidiary model. In the proposed adversarial process using cosine similarity, on the contrary, the performance of the subsidiary model can be degraded more efficiently by searching feature space orthogonal to the subsidiary model. The experiments on speaker identification and image recognition show that we found features that make the outputs of the subsidiary models independent of the input, and the performances of the primary models are improved.

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。

↑