Reducing the feature divergence of RGB and near-infrared images using Switchable Normalization
This work addresses multi-modality challenges in agricultural aerial image processing, but it is incremental as it builds on existing DeepLabV3 models with specific adaptations.
The paper tackled the problem of feature divergence between RGB and near-infrared images in agricultural aerial image segmentation, resulting in nearly 10% improvement in mean IoU over baselines by using Switchable Normalization and a hybrid loss function.
Visual pattern recognition over agricultural areas is an important application of aerial image processing. In this paper, we consider the multi-modality nature of agricultural aerial images and show that naively combining different modalities together without taking the feature divergence into account can lead to sub-optimal results. Thus, we apply a Switchable Normalization block to our DeepLabV3 segmentation model to alleviate the feature divergence. Using the popular symmetric Kullback Leibler divergence measure, we show that our model can greatly reduce the divergence between RGB and near-infrared channels. Together with a hybrid loss function, our model achieves nearly 10\% improvements in mean IoU over previously published baseline.