Feature Pyramid Network for Multi-Class Land Segmentation

TL;DR

This paper introduces a feature pyramid network-based approach for multi-class land segmentation in satellite imagery, demonstrating reliable results and efficient memory usage suitable for common GPU hardware.

Contribution

It presents a novel FPN-based neural network architecture with a ResNet50 encoder for improved multi-class land segmentation in satellite images.

Findings

Achieved reliable results on DEEPGLOBE land cover challenge

Network performs well with moderate memory requirements

Allows fast predictions using standard GPUs

Abstract

Semantic segmentation is in-demand in satellite imagery processing. Because of the complex environment, automatic categorization and segmentation of land cover is a challenging problem. Solving it can help to overcome many obstacles in urban planning, environmental engineering or natural landscape monitoring. In this paper, we propose an approach for automatic multi-class land segmentation based on a fully convolutional neural network of feature pyramid network (FPN) family. This network is consisted of pre-trained on ImageNet Resnet50 encoder and neatly developed decoder. Based on validation results, leaderboard score and our own experience this network shows reliable results for the DEEPGLOBE - CVPR 2018 land cover classification sub-challenge. Moreover, this network moderately uses memory that allows using GTX 1080 or 1080 TI video cards to perform whole training and makes pretty fast predictions.