Streaming Networks：通过硬连线和输入诱导稀疏性提高噪声鲁棒性和滤波器多样性
Streaming Networks: Increase Noise Robustness and Filter Diversity via Hard-wired and Input-induced Sparsity
摘要：
CNN已经在许多应用中实现了最先进的性能。最近的研究表明，如果图像被噪声破坏，CNN的识别准确率会急剧下降。我们专注于噪声污染的图像的鲁棒识别精度的问题。我们介绍了一种新的网络体系结构称为流网络。每个流都以原始图像的某个强度切片作为输入，并且流参数被独立地训练。我们使用网络容量，硬连线和输入引起的稀疏度作为实验的维度。结果表明，只有同时存在硬连线和输入诱导稀疏性，才能实现鲁棒的噪声图像识别。流网络是唯一的架构，同时具有两种类型的稀疏性，并表现出更高的鲁棒性噪声。最后，为了说明滤波器多样性的增加，我们说明了随着硬连线和域引起的稀疏度和容量的增加，第一卷积层的滤波器权重的分布逐渐接近均匀分布。
英文摘要：
The CNNs have achieved a state-of-the-art performance in many applications. Recent studies illustrate that CNN's recognition accuracy drops drastically if images are noise corrupted. We focus on the problem of robust recognition accuracy of noise-corrupted images. We introduce a novel network architecture called Streaming Networks. Each stream is taking a certain intensity slice of the original image as an input, and stream parameters are trained independently. We use network capacity, hard-wired and input-induced sparsity as the dimensions for experiments. The results indicate that only the presence of both hard-wired and input-induces sparsity enables robust noisy image recognition. Streaming Nets is the only architecture which has both types of sparsity and exhibits higher robustness to noise. Finally, to illustrate increase in filter diversity we illustrate that a distribution of filter weights of the first conv layer gradually approaches uniform distribution as the degree of hard-wired and domain-induced sparsity and capacities increases.
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。
