Large Pigment Database
<div> <div> <p><em>Nowadays, large language models are increasingly human-level intelligent, yet for primitive tasks such as sensing chemicals – dogs continue to surpass any man-made technology. A main reason for the underdevelopment of AI technology for chemosensing is the lack of a large and open-access dataset. As a step to fix this, we may create a large pigment model by curating already published open-access datasets containing the spectral reflectance properties of 1000s of pigments. This can lead to novel hyperspectral computer-vision technology that can detect the mixing ratios of pigments used in a painting, monitor the nutritional profile of crops, detect skin diseases etc. </em></p> <p>Your <strong>reflectance spectra datasets</strong><em> </em>are highly welcome because presently available datasets on spectrochemical analysis of pigments have the following limitations:</p> <p><a href="https://spectrabase.com/">SpectraBase</a> (belongs to a company that forbids open-access machine learning development)</p> <p><a href="https://www.sciencebase.gov/catalog/item/5807a2a2e4b0841e59e3a18d">USGS Spectral Library Version 7 Data - ScienceBase-Catalog</a> (contains some spectra with reflectance values above 1)</p> <p><a href="https://chsopensource.org/pigments-checker/">Pigments Checkers – Cultural Heritage Science Open Source (chsopensource.org)</a><strong> </strong>(contains less than 100 pigments)</p> <p><a href="https://www.nature.com/articles/nature14576">A colloidal quantum dot spectrometer | Nature</a><strong> </strong>(contains spectra plotted in images, needs conversion to an array of numbers)</p> <p><a href="https://soprano.kikirpa.be/">SOPRANO (kikirpa.be)</a><strong> </strong>(contains Raman spectra where the excitation frequency is not standardised)</p> <p><a href="https://www.fpbase.org/protein/dlanyfp/">dLanYFP :: Fluorescent Protein Database (fpbase.org)</a><strong> </strong>(contains Fluorescence spectra where the excitation frequency is not standardised)</p> </div> </div>
ShareScore
32/100
Overall dataset sharing score
Score breakdown
These five areas show where the dataset supports — or may limit — practical reuse.
- Stewardship
- 4
- Harmonization
- 4
- Access
- 16
- Reuse readiness
- 8
- Engagement
- 0
ScienceDex for faculty
Your academic CV and NIH biosketch
Make your CV in your school's format and your NIH biosketch from it, then download both as Word files.