A novel approach of data augmentation based on irregular superpixel decomposition is proposed. This approach called SuperpixelGridMasks permits to extend original image datasets that are required by training stages of machine learning-related analysis architectures towards increasing their performances.
These grid-based methods produce a new style of image transformations using the dropping and fusing of information.
Extensive experiments using various image classification models as well as precision health and surrounding real-world datasets show that baseline performances can be significantly outperformed using our methods. The comparative study also shows that our methods can overpass the performances of other data augmentations.
More than ever, machine learning is a field of great interest for the community of health informatics. To the past, a huge amount of approaches have been developed to perform semi-automatic and automatic data segmentation and classification towards supporting health diagnoses and people wellbeing.
In this context, deep convolutional neural networks CNNs appear as an ultimate resource towards operating recognition tasks at various scales. Indeed, recognition models are created by exploiting, e.
