Applying Knowledge Transfer for Water Body Segmentation in Peru
This addresses a practical remote sensing problem for environmental monitoring in Peru, but the results are negative/incremental.
The paper tackled water body segmentation in Peruvian satellite images using a knowledge transfer approach to address scarce labeled data, but found that incorporating high-resolution reference data actually worsened segmentation performance, suggesting distribution mismatch issues.
In this work, we present the application of convolutional neural networks for segmenting water bodies in satellite images. We first use a variant of the U-Net model to segment rivers and lakes from very high-resolution images from Peru. To circumvent the issue of scarce labelled data, we investigate the applicability of a knowledge transfer-based model that learns the mapping from high-resolution labelled images and combines it with the very high-resolution mapping so that better segmentation can be achieved. We train this model in a single process, end-to-end. Our preliminary results show that adding the information from the available high-resolution images does not help out-of-the-box, and in fact worsen results. This leads us to infer that the high-resolution data could be from a different distribution, and its addition leads to increased variance in our results.