Spectroscopic sensing
Cross-device Raman spectroscopy
Can spectra recorded on different instruments support more consistent recognition of the same analyte classes?
Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy · 2026
Approach
SERS-D2DNet transforms spectra from four portable devices toward a Renishaw reference. The study then uses SuperRaman to classify the original and transformed spectra across 20 analyte classes, using 50 spectra per class to train the classifiers.
Reported results
Classification accuracy increased after transformation for all four portable devices in Table 5. The largest gains were for Enwave and Rapid ID.
- Enwave accuracy
- 97.58% → 99.98%
- Rapid ID accuracy
- 96.54% → 98.87%
My contribution
I led the modelling and evaluation of the spectral transformation and recognition workflow, including implementation, generalization analysis, visualization, and writing.
Study scope
These comparisons cover the analytes and acquisition conditions tested in the study. The Zenodo record describes the dataset and how to request access to the restricted files.
