Trigonometric-Euclidean-Smoother Interpolator (TESI) for continuous time-series and non-time-series data augmentation for deep neural network applications in agriculture

•A new method is proposed for data augmentation for deep neural network use.•The method uses a trigonometric-Euclidian space to generate the new data points.•The new method is compared to the deep learning-based augmentation methods.•The new method retained the data's original distribution, gai...

Full description