Probabilistic Deep Discriminant Analysis for Wind Blade Segmentation

TL;DR

This paper introduces Probabilistic Deep Discriminant Analysis (PDDA), a novel method that enhances wind blade segmentation by optimizing class separation with deep networks, improving accuracy and confidence in predictions.

Contribution

The paper presents the first application of Deep Discriminant Analysis to image segmentation, developing stable loss functions and a probabilistic framework for improved class separation.

Findings

PDDA achieves higher segmentation accuracy.

PDDA produces more confident and consistent predictions.

Significant reduction in within-class variance.

Abstract

Linear discriminant analysis improves class separability but struggles with non-linearly separable data. To overcome this, we introduce Deep Discriminant Analysis (DDA), which directly optimizes the Fisher criterion utilizing deep networks. To ensure stable training and avoid computational instabilities, we incorporate signed between-class variance, bound outputs with a sigmoid function, and convert multiplicative relationships into additive ones. We present two stable DDA loss functions and augment them with a probability loss, resulting in Probabilistic DDA (PDDA). PDDA effectively minimizes class overlap in output distributions, producing highly confident predictions with reduced within-class variance. When applied to wind blade segmentation, PDDA showcases notable advances in performance and consistency, critical for wind energy maintenance. To our knowledge, this is the first application of DDA to image segmentation.