Exploiting Convolutional Representations for Multiscale Human Settlement Detection

TL;DR

This paper proposes a unified framework for extracting and visualizing features from a single deep convolutional network to improve multiscale human settlement detection in remote sensing images without retraining.

Contribution

It introduces a novel approach to reuse trained convolutional features across multiple remote sensing tasks, demonstrating potential for transfer learning and multiscale analysis.

Findings

Unified feature extraction framework for multiple tasks

Effective visualization of different image representation spaces

Preliminary evidence of transfer learning potential

Abstract

We test this premise and explore representation spaces from a single deep convolutional network and their visualization to argue for a novel unified feature extraction framework. The objective is to utilize and re-purpose trained feature extractors without the need for network retraining on three remote sensing tasks i.e. superpixel mapping, pixel-level segmentation and semantic based image visualization. By leveraging the same convolutional feature extractors and viewing them as visual information extractors that encode different image representation spaces, we demonstrate a preliminary inductive transfer learning potential on multiscale experiments that incorporate edge-level details up to semantic-level information.