We just released v0.2 of tsaug, an open-source Python package for data augmentation of time series and audio sequences.
Data augmentation is a common technique applied in deep learning. Augmentation of image data for computer vision models has been a mature and standard process, and several open-source packages (e.g. imgaug) are popular choices. However, open-source libraries for data augmentation of time series (including audio sequences) are rare.
In tsaug, we developed several common augmenters of time series and audio sequences as well as pipelines that may connect augmenter instances. More importantly, tsaug is an expandable platform where new augmenters might be developed following a unified API.
Issues and PR are welcomed, as well as stars and forks!
An example of time series: https://arundo-tsaug.readthedocs-hosted.com/en/stable/quicks...
An example of audio sequences: https://arundo-tsaug.readthedocs-hosted.com/en/stable/quicks...
