Graphene-PbS photodetector platform for smart contact lenses data

The dataset consists of the raw data used to train a variational autoencoder on classifying blinks acquired from a graphene-PbS quantum dot photodetector mounted on a wearable contact lens platform, simulation and measurement data for the NFC coil used on this platform and raw readings at low illuminance ranges. Most data can be used as is. However, to train the autoencoder, no stimuli containing regions from the files in folder 'ML training data' are stitched together, normalised and then segmented into timeframes of length 81 to create the training set. A similar normalisation procedure is applied to the test data (files in the folder ML Low Illuminance Test Data).