SPIN: Simplifying Polar Invariance for Neural networks Application to vision-based irradiance forecasting
This work addresses a domain-specific problem in solar energy forecasting by simplifying rotational invariance, offering incremental improvements over existing methods.
The paper tackles the challenge of rotational invariance in convolutional neural networks for vision-based solar irradiance forecasting by converting images to polar coordinates, which transforms rotational invariance into translational invariance. This approach improves prediction results and reduces training time by a factor of 4 compared to data augmentation with rotations.
Translational invariance induced by pooling operations is an inherent property of convolutional neural networks, which facilitates numerous computer vision tasks such as classification. Yet to leverage rotational invariant tasks, convolutional architectures require specific rotational invariant layers or extensive data augmentation to learn from diverse rotated versions of a given spatial configuration. Unwrapping the image into its polar coordinates provides a more explicit representation to train a convolutional architecture as the rotational invariance becomes translational, hence the visually distinct but otherwise equivalent rotated versions of a given scene can be learnt from a single image. We show with two common vision-based solar irradiance forecasting challenges (i.e. using ground-taken sky images or satellite images), that this preprocessing step significantly improves prediction results by standardising the scene representation, while decreasing training time by a factor of 4 compared to augmenting data with rotations. In addition, this transformation magnifies the area surrounding the centre of the rotation, leading to more accurate short-term irradiance predictions.