Volatile Organic Compounds (VOCs) are organic molecules that have low boiling points and therefore easily evaporate in the air. They pose significant risks to human health, making their accurate detection crucial for efforts to monitor and minimize exposure. Infrared (IR) spectroscopy enables the ultrasensitive detection of VOCs at low-concentrations in the atmosphere by measuring their IR absorption spectra. However, the complexity of the IR spectra limits the possibility to implement VOC recognition and quantification in real-time. While deep neural networks (NNs) are increasingly used for the recognition of complex data structures, they typically require massive datasets for the training phase. Here, we create an experimental VOC dataset for nine different classes of compounds at various concentrations, using their IR absorption spectra. To further increase the amount of spectra and their diversity in terms of VOC concentration, we augment the experimental dataset with synthetic spectra created via conditional generative NNs. This approach allows us to train robust discriminative NNs, able to reliably identify the nine VOCs, as well as to precisely predict their concentrations. The trained NN is suitable for integration into sensing devices for VOCs recognition and analysis. • IR spectroscopy is applied for recognition and quantification of volatile organic compounds. • Deep convolutional neural networks are able to discriminate and quantify volatile organic compounds from IR absorption spectra. • Conditional variation autoencoder can be used to produce synthetic spectra to perform data augmentation and increase the diversity of the dataset.
Valle et al. (Thu,) studied this question.