EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Generative Artificial Intelligence for Synthetic Spectral Data Augmentation in Sensor-Based Plastic Recycling

Roman-David Kulko iD, Andreas Hanus iD, Benedikt Elser iD

Sensors · 01.07.2025 · 25 (13) · 4114

DOI: 10.3390/s25134114 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Sensors, online erschienen am 01.07.2025. Verfasst von Roman-David Kulko und 2 weiteren Autor:innen. Beteiligte Einrichtung: Technologie Campus Grafenau, Technische Hochschule Deggendorf, 94481 Grafenau, Germany. Die Arbeit wurde bisher 7-mal zitiert (Crossref) und nennt 40 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Bavarian Ministry of Economic Affairs, Re- gional Development and Energy, Bavarian Ministry of Economic Affairs, Regional Development and Energy.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The reliance on deep learning models for sensor-based material classification amplifies the demand for labeled training data. However, acquiring large-scale, annotated spectral data for applications such as near-infrared (NIR) reflectance spectroscopy in plastic sorting remains a significant challenge due to high acquisition costs and environmental variability. This paper investigates the potential of large language models (LLMs) in synthetic spectral data generation. Specifically, it examines whether LLMs have acquired sufficient implicit knowledge to assist in generating spectral data and introduce meaningful variations that enhance model performance when used for data augmentation. Classification accuracy is reported exclusively as a proxy for structural plausibility of the augmented spectra; maximizing augmentation performance itself is not the study’s goal. From as little as one empirical mean spectrum per class, LLM-guided simulation produced data that enabled up to 86% accuracy, evidence that the generated variation preserves class-distinguishing information. While the approach performs best for spectral distinct polymers, overlapping classes remain challenging. Additionally, the transfer of optimized augmentation parameters to unseen classes indicates potential for generalization across material types. While plastic sorting serves as a case study, the methodology may be applicable to other domains such as agriculture or food quality assessment, where spectral data are limited. The study outlines a novel path toward scalable, AI-supported data augmentation in spectroscopy-based classification systems.

Schlagwörter

Stichwörter der Autor:innen

deep learning for recyclinggenerative artificial intelligenceindustrial material classificationlarge language modelsnear-infrared spectroscopyplastic waste sortingspectral data augmentationsynthetic data generation

Autor:innen und Institutionen

  • Technologie Campus Grafenau, Technische Hochschule Deggendorf, 94481 Grafenau, Germany.
  • Sesotec GmbH, Regener Straße 130, 94513 Schönberg, Germany.

ORCID-Profile: Roman-David Kulko, Andreas Hanus, Benedikt Elser

Förderung

  • Bavarian Ministry of Economic Affairs, Re- gional Development and Energy
  • Bavarian Ministry of Economic Affairs, Regional Development and Energy

Zitieren

APA

Kulko, R. D., Hanus, A., & Elser, B. (2025). Generative Artificial Intelligence for Synthetic Spectral Data Augmentation in Sensor-Based Plastic Recycling. Sensors, 25(13), 4114. https://doi.org/10.3390/s25134114

Vancouver

Kulko RD, Hanus A, Elser B. Generative Artificial Intelligence for Synthetic Spectral Data Augmentation in Sensor-Based Plastic Recycling. Sensors. 2025;25(13):4114. doi:10.3390/s25134114

BibTeX
@article{Kulko2025,
  author = {Kulko, Roman-David and Hanus, Andreas and Elser, Benedikt},
  title = {Generative Artificial Intelligence for Synthetic Spectral Data Augmentation in Sensor-Based Plastic Recycling},
  journal = {Sensors},
  year = {2025},
  volume = {25},
  number = {13},
  pages = {4114},
  doi = {10.3390/s25134114}
}
Kostenlos starten →