针对声学场景分类的神经结构搜索（CS CompLang）
- 2020 年 1 月 2 日
- 笔记
卷积神经网络在声学场景分类（ASC）任务中被广泛采用，但是它们通常会带来沉重的计算负担。在这项工作中，我们提出了一个受MobileNetV2启发的轻量级高性能基线网络，该网络用单向内核替换了正方形卷积内核，在时间和频率维度上交替提取特征。此外，我们使用最新的神经架构搜索（NAS）范式探索了在建议的基准基础上构建的动态架构空间，该范式首先训练一个包含所有候选网络的超网，然后将著名的进化算法NSGA-II用于发现效率更高，精度更高且计算成本更低的网络。实验结果表明，我们搜索到的网络能够胜任ASC任务，在DCASE2018任务5评估集上达到90.3％的F1得分，这标志着一种新的先进性能，同时与我们的基准网络相比节省了25％的FLOP 。
原文题目：Neural Architecture Search on Acoustic Scene Classification
原文：Convolutional neural networks are widely adopted in Acoustic Scene Classification (ASC) tasks, but they generally carry a heavy computational burden. In this work, we propose a lightweight yet high-performing baseline network inspired by MobileNetV2, which replaces square convolutional kernels with unidirectional ones to extract features alternately in temporal and frequency dimensions. Furthermore, we explore a dynamic architecture space built on the basis of the proposed baseline with the recent Neural Architecture Search (NAS) paradigm, which first trains a supernet that incorporates all candidate networks and then applies a well-known evolutionary algorithm NSGA-II to discover more efficient networks with higher accuracy and lower computational cost. Experimental results demonstrate that our searched network is competent in ASC tasks, which achieves 90.3% F1-score on the DCASE2018 task 5 evaluation set, marking a new state-of-the-art performance while saving 25% of FLOPs compared to our baseline network.
原文作者：Jixiang Li,Chuming Liang,Bo Zhang,Zhao Wang,Fei Xiang,Xiangxiang Chu
原文地址：https://arxiv.org/abs/1912.12825
针对声学场景分类的神经结构搜索（CS CompLang）.pdf
