Data augmentation is an important term in the fields of artificial intelligence, big data, and smart data, as well as automation. It refers to methods by which existing datasets are artificially enlarged or expanded to create a better basis for the development and training of algorithms.
Suppose a company wants to develop an artificial intelligence that can automatically recognise product photos. For this, the AI needs many different images of each product. However, there are often not enough photos available. This is where data augmentation comes in: Computer programs are used to create new variations from existing images – for example, by changing the colours, mirroring the image, or rotating it slightly. This quickly creates many additional data points without much effort.
The advantage of data augmentation is that algorithms become more robust and accurate because they can be trained on a greater variety of examples. This saves time and money because less real data needs to be collected, and it significantly improves the quality of Artificial Intelligence results.
Deutsch 
English (UK)