密集卷积神经网络用于识别拉曼光谱
Wei Zhou1, Ziheng Qian1, Xinyuan Ni1
1Engineering Research Center of Optical Instrument and System, Ministry of Education, Shanghai Key Laboratory of Modern Optical System, School of Optical-Electrical and Computer Engineering, University of Shanghai for Science and Technology, 516 Jungong Rd., Shanghai 200093, China.
使用密集卷积神经网络的新型深度学习算法增强了基于云的拉曼光谱识别. 这种方法即使在环境干扰的情况下也能达到高精度,从而改善化学物质的检测.
科学领域:
- 频谱学是一种光谱学.
- 人工智能的人工智能
- 云计算 云计算 云计算
背景情况:
- 云计算和深度学习使智能应用得到广泛应用.
- 拉曼光谱识别可以在云中执行,减少对终端仪器的依赖.
- 环境干扰可以显著降低拉曼光谱识别算法的准确性.
研究的目的:
- 为准确的基于云的拉曼光谱识别提出一个深度学习算法.
- 为应对因环境干扰而导致识别精度下降的挑战.
- 提高拉曼光谱识别在各种条件下的稳定性和适应性.
主要方法:
- 开发了一个基于超过40个层的密集卷积神经网络 (CNN) 的深度学习算法.
- 密集网络在其密集块中具有前连接,减轻梯度问题并增强功能重用.
- 为了测试,创建了一个来自32种液体化学品的1600个拉曼光谱的数据库,包括干扰光谱.
主要成果:
- 拟议的Dense CNN与其他基于CNN的算法相比,表现出卓越的准确性和稳定性.
- 该算法在50个重复训练和测试集中在定制数据库上实现了99.99%的加权准确率.
- 在RRUFF数据库上测试时,Dense网络也表现良好.
结论:
- 开发的基于密集网络的方法显著推进了云支持的拉曼光谱识别.
- 该方法有效地减轻噪声干扰,即使在复杂的环境中也能确保精确的识别.
- 这种方法为各种拉曼光谱识别任务提供了更好的准确性和适应性.
更多相关视频
07:37An Integrated Raman Spectroscopy and Mass Spectrometry Platform to Study Single-Cell Drug Uptake, Metabolism, and Effects
Published on: January 9, 2020
07:51Combining Raman Imaging and Multivariate Analysis to Visualize Lignin, Cellulose, and Hemicellulose in the Plant Cell Wall
Published on: June 10, 2017
相关概念视频
Raman Spectroscopy: Overview
However, a small fraction of the scattered light exhibits a frequency shift due to the exchange of energy between the incident photons and...
Raman Spectroscopy Instrumentation: Overview
The monochromatic laser source, typically using visible or near-infrared radiation, generates a highly focused beam of light. This light interacts with the molecules of the sample, scattering some of the light. Liquid and gaseous samples are usually tested in ordinary glass capillaries, while solids can be analyzed as powders packed in capillaries or as potassium...
Methods of Classification and Identification
IR Frequency Region: Fingerprint Region
