Crop Classification under Varying Cloud Cover with Neural Ordinary Differential Equations

TL;DR

This paper introduces an ODE-RNN model combining neural ordinary differential equations with recurrent neural networks to improve crop classification accuracy from irregularly sampled satellite image sequences, especially under cloud cover.

Contribution

It presents a novel ODE-RNN approach that models continuous latent dynamics, enabling better handling of irregular sampling and early-season forecasting in crop classification.

Findings

ODE-RNN outperforms LSTM, GRU, and temporal convolution baselines.

Significant accuracy gains in scenarios with few observations.

Enhanced early-season classification through better extrapolation.

Abstract

Optical satellite sensors cannot see the Earth's surface through clouds. Despite the periodic revisit cycle, image sequences acquired by Earth observation satellites are therefore irregularly sampled in time. State-of-the-art methods for crop classification (and other time series analysis tasks) rely on techniques that implicitly assume regular temporal spacing between observations, such as recurrent neural networks (RNNs). We propose to use neural ordinary differential equations (NODEs) in combination with RNNs to classify crop types in irregularly spaced image sequences. The resulting ODE-RNN models consist of two steps: an update step, where a recurrent unit assimilates new input data into the model's hidden state; and a prediction step, in which NODE propagates the hidden state until the next observation arrives. The prediction step is based on a continuous representation of the latent dynamics, which has several advantages. At the conceptual level, it is a more natural way to describe the mechanisms that govern the phenological cycle. From a practical point of view, it makes it possible to sample the system state at arbitrary points in time, such that one can integrate observations whenever they are available, and extrapolate beyond the last observation. Our experiments show that ODE-RNN indeed improves classification accuracy over common baselines such as LSTM, GRU, and temporal convolution. The gains are most prominent in the challenging scenario where only few observations are available (i.e., frequent cloud cover). Moreover, we show that the ability to extrapolate translates to better classification performance early in the season, which is important for forecasting.