Scientific purpose and scope 

The TEM Nanoparticle segmentation database (TEM-ParticlesDB) is a free repository of experimental Transmission Electron Microscopy (TEM) images of Nanoparticle which have been accurately segmented to identify size, shape and morphology of the particles observed, across a range of substrates. 

Each record can contain experimental data, synthetic data and segmented nanoparticle data. 

Using the Transmission Electron Microscopy Particle Outline Segmentation (TEMPOS) software, computationally synthesised data can be generated to train a segmentation model via transfer learning. This model is then applied to segment experimental data. Both the synthetic training data and the resulting segmentation outputs are archived and made available through TEM-ParticlesDB. 

We anticipate that the next generation of machine learning models will benefit from the availability of these experimental, synthetic, and segmentation datasets, facilitating the development of increasingly robust algorithms for automated nanoscale object detection in TEM images. 

We therefore encourage researchers to share their data from their published, and unpublished work to benefit the wider research community. 

This collection is part of the “AI for Science” PSDI sub-project (“Provision of ‘AI ready’ data: prototyping data pipelines and repositories”, grant application APP84520, award UKRI2697, opportunity OPP1033:EPSRC AI for Science) and as such has been selected and prepared as an example of “AI ready” data that is useful as training data for AI tools and systems.  

Any feedback, suggestions or contributions can be provided to support@psdi.ac.uk. Specifically, feedback regarding the use this dataset by AI tools and systems would be particularly welcome.