Land cover mapping at very high resolution with rotation equivariant CNNs: towards small yet accurate models

Abstract

In remote sensing images, the absolute orientation of objects is arbitrary. Depending on an object's orientation and on a sensor's flight path, objects of the same semantic class can be observed in different orientations in the same image. Equivariance to rotation, in this context understood as responding with a rotated semantic label map when subject to a rotation of the input image, is therefore a very desirable feature, in particular for high capacity models, such as Convolutional Neural Networks (CNNs). If rotation equivariance is encoded in the network, the model is confronted with a simpler task and does not need to learn specific (and redundant) weights to address rotated versions of the same object class. In this work we propose a CNN architecture called Rotation Equivariant Vector Field Network (RotEqNet) to encode rotation equivariance in the network itself. By using rotating convolutions as building blocks and passing only the the values corresponding to the maximally activating orientation throughout the network in the form of orientation encoding vector fields, RotEqNet treats rotated versions of the same object with the same filter bank and therefore achieves state-of-the-art performances even when using very small architectures trained from scratch. We test RotEqNet in two challenging sub-decimeter resolution semantic labeling problems, and show that we can perform better than a standard CNN while requiring one order of magnitude less parameters.

References in corpus (10)

• Fully Convolutional Networks for Semantic Segmentation
• Land Use Classification in Remote Sensing Images by Convolutional Neural Networks
• Dense semantic labeling of sub-decimeter resolution images with convolutional neural networks
• Land cover mapping at very high resolution with rotation equivariant CNNs: towards small yet accurate models
• Fully Convolutional Networks for Dense Semantic Labelling of High-Resolution Aerial Imagery
• Rotation equivariant vector field networks
• Multiclass feature learning for hyperspectral image classification: sparse and hierarchical solutions
• Oriented Response Networks
• Harmonic Networks: Deep Translation and Rotation Equivariance
• TI-POOLING: transformation-invariant pooling for feature learning in Convolutional Neural Networks

Cited by in corpus (21)

• UNetFormer: A UNet-like Transformer for Efficient Semantic Segmentation of Remote Sensing Urban Scene Imagery
• Land cover mapping at very high resolution with rotation equivariant CNNs: towards small yet accurate models
• Recurrently Exploring Class-wise Attention in A Hybrid Convolutional and Bidirectional LSTM Network for Multi-label Aerial Image Classification
• iSAID: A Large-scale Dataset for Instance Segmentation in Aerial Images
• LoveDA: A Remote Sensing Land-Cover Dataset for Domain Adaptive Semantic Segmentation
• Aerial Imagery for Roof Segmentation: A Large-Scale Dataset towards Automatic Mapping of Buildings
• Enabling Country-Scale Land Cover Mapping with Meter-Resolution Satellite Imagery
• Relation Network for Multi-label Aerial Image Classification
• Deep multi-task learning for a geographically-regularized semantic segmentation of aerial images
• Building Footprint Generation by IntegratingConvolution Neural Network with Feature PairwiseConditional Random Field (FPCRF)
• Correcting rural building annotations in OpenStreetMap using convolutional neural networks
• Semantic Labeling of High Resolution Images Using EfficientUNets and Transformers
• RSI-Net: Two-Stream Deep Neural Network for Remote Sensing Imagesbased Semantic Segmentation
• A large-scale climate-aware satellite image dataset for domain adaptive land-cover semantic segmentation
• MultiScene: A Large-scale Dataset and Benchmark for Multi-scene Recognition in Single Aerial Images
• Hybrid Multiple Attention Network for Semantic Segmentation in Aerial Images
• A Relation-Augmented Fully Convolutional Network for Semantic Segmentation in Aerial Scenes
• Transformer Meets Convolution: A Bilateral Awareness Network for Semantic Segmentation of Very Fine Resolution Urban Scene Images
• PointFlow: Flowing Semantics Through Points for Aerial Image Segmentation
• A deep network approach to multitemporal cloud detection
• Aerial Scene Understanding in The Wild: Multi-Scene Recognition via Prototype-based Memory Networks