Extending the spectral database of laser-induced breakdown spectroscopy with generative adversarial nets

Teng, Geer; Wang, Q. Q.; Kong, Jinglin; Dong, Liqiang; Cui, Xutai; Liu, W. W.; Wei, Kai; Xiangli, Wenting

doi:10.1364/oe.27.006958

Cited by 34 publications

(10 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A generative adversarial network (GAN)-based semisupervised method can provide accurate classification with far fewer labeled samples than CNN-based supervised methods (Wang et al, 2017). Teng et al (2019) found that a GAN-based spectral generation method can be effectively used to extend a spectral database. When PCA and the K-means algorithm were used for clustering evaluation, it was found that the images generated by the GAN were classified as real images.…”

Section: Introductionmentioning

confidence: 99%

Classification of pathogens by Raman spectroscopy combined with generative adversarial networks

et al. 2020

Science of The Total Environment

View full text Add to dashboard Cite

Section: Introductionmentioning

confidence: 99%

Classification of pathogens by Raman spectroscopy combined with generative adversarial networks

et al. 2020

Science of The Total Environment

View full text Add to dashboard Cite

“…PCA is a widely used algorithm for dimensionality reduction and feature extraction of Vis–NIR data [ 62 ]. According to studies by Sun et al [ 22 ] and Teng et al [ 63 ], the dimensionality reduction of spectral data using PCA allows the similarity and diversity of the generated spectra compared to the original spectra to be assessed. The basic principle of PCA is to convert the original m-dimensional data into a new set of k-dimensional orthogonal variables (called principal components, PCs).…”

Section: Methodsmentioning

confidence: 99%

Vis–NIR Spectroscopy Combined with GAN Data Augmentation for Predicting Soil Nutrients in Degraded Alpine Meadows on the Qinghai–Tibet Plateau

Jiang

Zhao

Ding

et al. 2023

Sensors

View full text Add to dashboard Cite

Soil nutrients play vital roles in vegetation growth and are a key indicator of land degradation. Accurate, rapid, and non-destructive measurement of the soil nutrient content is important for ecological conservation, degradation monitoring, and precision farming. Currently, visible and near-infrared (Vis–NIR) spectroscopy allows for rapid and non-destructive monitoring of soil nutrients. However, the performance of Vis–NIR inversion models is extremely dependent on the number of samples. Limited samples may lead to low prediction accuracy of the models. Therefore, modeling and prediction based on a small sample size remain a challenge. This study proposes a method for the simultaneous augmentation of soil spectral and nutrient data (total nitrogen (TN), soil organic matter (SOM), total potassium oxide (TK2O), and total phosphorus pentoxide (TP2O5)) using a generative adversarial network (GAN). The sample augmentation range and the level of accuracy improvement were also analyzed. First, 42 soil samples were collected from the pika disturbance area on the QTP. The collected soils were measured in the laboratory for Vis–NIR and TN, SOM, TK2O, and TP2O5 data. A GAN was then used to augment the soil spectral and nutrient data simultaneously. Finally, the effect of adding different numbers of generative samples to the training set on the predictive performance of a convolutional neural network (CNN) was analyzed and compared with another data augmentation method (extended multiplicative signal augmentation, EMSA). The results showed that a GAN can generate data very similar to real data and with better diversity. A total of 15, 30, 60, 120, and 240 generative samples (GAN and EMSA) were randomly selected from 300 generative samples to be included in the real data to train the CNN model. The model performance first improved and then deteriorated, and the GAN was more effective than EMSA. Further shortening the interval for adding GAN data revealed that the optimal ranges were 30–40, 50–60, 30–35, and 25–35 for TK2O, TN, TP2O5, and SOM, respectively, and the validation set accuracy was maximized in these ranges. Therefore, the above method can compensate to some extent for insufficient samples in the hyperspectral prediction of soil nutrients, and can quickly and accurately estimate the content of soil TK2O, TN, TP2O5, and SOM.

show abstract

“…This is because these parameters affect the LIBS plasma for standoff and trace analysis of any hazardous materials. Additionally, developing superior semi-supervised/supervised and/or machine learning based algorithms will help in not only distinguishing the explosives but also for unambiguous identification and labelling [54][55] .…”

Section: Conclusion and Future Scopementioning

confidence: 99%

Femtosecond Filaments for Standoff Detection of Explosives

Soma

Shaik

2020

Def. Sc. Jl.

View full text Add to dashboard Cite

In this report, we present our results from various studies to qualitatively discriminate the common military explosives viz. RDX, TNT and HMX in their pure form at a distance of ~6.5 m in standoff mode using femtosecond (fs) filament induced breakdown spectroscopy technique (fs FIBS) together with principal component analysis. A ~30 cm length fs filament obtained by a two-lens configuration was used to interrogate those energetic molecules in the form of pressed pellets (150 mg each). The plasma emissions were collected by a Schmidt-Cassegrain telescope (SCT) from a distance of ~8 m away from the investigation zone. Additionally, a few significant results obtained from the LIBS-based investigations of nitroimidazoles with respect to the standoff distance (~2 m) are discussed. Furthermore, we have also summarised a few important results from our recent investigations of bulk energetic materials in various configurations (including those with fs filaments). The results obtained from various fs FIBS configurations corroborate that the filament generation and its properties, the size and f-number of collection optics influence signal strength in the FIBS technique. These results project the fs FIBS technique as a potential technique for investigations aimed at hazardous materials and harsh environments in the standoff mode.

show abstract

Extending the spectral database of laser-induced breakdown spectroscopy with generative adversarial nets

Cited by 34 publications

References 29 publications

Classification of pathogens by Raman spectroscopy combined with generative adversarial networks

Classification of pathogens by Raman spectroscopy combined with generative adversarial networks

Vis–NIR Spectroscopy Combined with GAN Data Augmentation for Predicting Soil Nutrients in Degraded Alpine Meadows on the Qinghai–Tibet Plateau

Femtosecond Filaments for Standoff Detection of Explosives

Contact Info

Product

Resources

About