Abstract:This contribution provides an interdisciplinary perspective on current research and technological developments associated with satellite remote sensing and deep learning that have the potential to make a difference in wildlife management. We highlight current barriers to the broader adoption of these approaches by experts operating at the interface between satellite remote sensing and ecology and conservation and discuss how these can be overcome. In the context of the current nature crisis, being able to reliably and cost‐effectively track subtle changes in the biosphere across adequate spatial and temporal extents and resolutions is crucial. Deep learning represents a group of versatile approaches to image processing tasks that are increasingly combined with satellite remote sensing imagery to monitor biodiversity and inform ecology and conservation, yet an overview of the opportunities and challenges associated with this development has so far been lacking. Here, we provide an interdisciplinary perspective on current research and technological developments associated with satellite remote sensing and deep learning that have the potential to make a difference in biodiversity monitoring and wildlife conservation; highlight challenges to the broader adoption of these approaches by experts operating at the interface between satellite remote sensing and ecology and conservation; and discuss how these can be overcome. By enabling the leveraging of big data and by providing new ways to learn about biodiversity and its dynamics, deep learning approaches promise to become a powerful tool to help address current monitoring needs and knowledge gaps. In certain situations, deep learning approaches may moreover substantially reduce the time and resources required to analyse satellite imagery. However, issues relating to capacity building, reference data access, environmental costs as well as model interpretability, robustness and alignment need to be addressed to successfully capitalize on these opportunities.

Game Theoretic Training Enabled Deep Learning Solutions for Rapid Discovery of Satellite Behaviors

Satellite Attitude Tracking Decision Method based on Deep Deterministic Policy Gradient for Moving Target Observation

Deep Kalman-based Trajectory Estimation of Moving Target from Satellite Images

Satellite Attitude Tracking Control of Moving Targets Combining Deep Reinforcement Learning and Predefined-time Stability Considering Energy Optimization

Satellite Image Classification with Deep Learning

Deep Learning for Understanding Satellite Imagery: An Experimental Survey

Deep Learning Methods for Space Situational Awareness in Mega-Constellations Satellite-Based Internet of Things Networks

Strategy for on-orbit space object classification using deep learning

Few-shot satellite image classification for bringing deep learning on board OPS-SAT

SatelliteChangeNet: Deep Learning approach for Detection & Prediction

Space target recognition based on deep learning

Autonomous measurement and semantic segmentation of non-cooperative targets with deep convolutional neural networks

Deep Learning for Satellite Image Time Series Analysis: A Review

Deep learning and satellite remote sensing for biodiversity monitoring and conservation

Using Game Engines and Machine Learning to Create Synthetic Satellite Imagery for a Tabletop Verification Exercise

Sensing for Space Safety and Sustainability: A Deep Learning Approach with Vision Transformers

Focusing on the Big Picture: Insights into a Systems Approach to Deep Learning for Satellite Imagery

Deep-learning-based solution for data deficient satellite image segmentation

A game theoretic self-organization for satellite-based optical sensor allocation

Model Optimization for Deep Space Exploration via Simulators and Deep Learning

Satellite Image Based Animal Identification System Using Deep Learning Assisted Remote Sensing Strategy