Polarizing Front Ends for Robust Cnns

机译：极化前端，可实现可靠的CNN

获取原文

页面导航

摘要
著录项
相似文献
相关主题

摘要

The vulnerability of deep neural networks to small, adversarially designed perturbations can be attributed to their "excessive linearity." In this paper, we propose a bottom-up strategy for attenuating adversarial perturbations using a nonlinear front end which polarizes and quantizes the data. We observe that ideal polarization can be utilized to completely eliminate perturbations, develop algorithms to learn approximately polarizing bases for data, and investigate the effectiveness of the proposed strategy on the MNIST and Fashion MNIST datasets.

机译：深度神经网络对较小的，经过对抗性设计的扰动的脆弱性可以归因于它们的“过度线性”。在本文中，我们提出了一种使用非线性前端对数据进行极化和量化的，用于衰减对抗性扰动的自下而上的策略。我们观察到理想极化可以用来完全消除干扰，开发算法以学习近似的极化数据基础，并研究MNIST和Fashion MNIST数据集上所提出策略的有效性。

著录项

来源
《IEEE International Conference on Acoustics, Speech and Signal Processing》|2020年|4257-4261|共5页
会议地点
作者
Can Bakiskan; Soorya Gopalakrishnan; Metehan Cekic; Upamanyu Madhow; Ramtin Pedarsani;
展开▼
作者单位

展开▼
会议组织
原文格式 PDF
正文语种
中图分类
关键词
adversarial machine learning; quantization; front-end defense;

机译：对抗机器学习;量化;前端防御;

相似文献

外文文献
中文文献
专利

1. A Robust Speaking Rate Estimator Using a CNN-BLSTM Network [J] . Srinivasan Aparna, Singh Diviya, Yarra Chiranjeevi, Circuits, systems and signal processing . 2021,第12期

机译：使用CNN-BLSTM网络强大的讲话速率估算器
2. Diagnosis of Alzheimer’s Disease Severity with fMRI Images Using Robust Multitask Feature Extraction Method and Convolutional Neural Network (CNN) [J] . Morteza Amini, Mir Mohsen Pedram, AliReza Moradi, Computational and mathematical methods in medicine . 2021,第a期

机译：使用鲁棒多任务的FMRI图像诊断阿尔茨海默病严重程度特征提取方法和卷积神经网络（CNN）
3. Robust regression with deep CNNs for facial age estimation: An empirical study [J] . Dornaika F., Bekhouche S. E., Arganda-Carreras I Expert Systems with Application . 2020,第Mara期

机译：深度CNN的稳健回归用于面部年龄估计：一项实证研究
4. Polarizing Front Ends for Robust Cnns [C] . Can Bakiskan, Soorya Gopalakrishnan, Metehan Cekic, IEEE International Conference on Acoustics, Speech and Signal Processing . 2020

机译：偏振前端为稳健的CNNS
5. Robust Acoustic Modeling and Front-End Design for Distant Speech Recognition [D] . Mirsamadi, Seyedmahdad. 2017

机译：鲁棒的声学建模和远端语音识别前端设计
6. Diagnosis of Alzheimers Disease Severity with fMRI Images Using Robust Multitask Feature Extraction Method and Convolutional Neural Network (CNN) [O] . Morteza Amini, MirMohsen Pedram, AliReza Moradi, 2021

机译：使用鲁棒多任务的FMRI图像诊断阿尔茨海默病严重程度特征提取方法和卷积神经网络（CNN）
7. Simulating a Primary Visual Cortex at the Front of CNNs Improves Robustness to Image Perturbations [O] . Joel Dapello, Tiago Marques, Martin Schrimpf, 2020

机译：模拟CNNS前面的主视觉皮质改善了图像扰动的鲁棒性

Polarizing Front Ends for Robust Cnns

摘要

著录项

相似文献

相关主题

期刊订阅