Deep Space
Deep space research encompasses a broad range of scientific inquiries, primarily focused on understanding and utilizing vast datasets from various sources, including satellite imagery, sensor readings, and astronaut photography. Current research heavily employs machine learning, particularly convolutional neural networks and transformers, to analyze these datasets for tasks like object detection, classification, and regression, often within specialized embedding spaces to improve efficiency and accuracy. These advancements are crucial for improving Earth observation, enabling more effective climate change mitigation (e.g., methane detection), optimizing resource management (e.g., aquaculture monitoring), and enhancing the efficiency and safety of space exploration.
Papers
Monitoring Vegetation From Space at Extremely Fine Resolutions via Coarsely-Supervised Smooth U-Net
Joshua Fan, Di Chen, Jiaming Wen, Ying Sun, Carla P. Gomes
CA-SpaceNet: Counterfactual Analysis for 6D Pose Estimation in Space
Shunli Wang, Shuaibing Wang, Bo Jiao, Dingkang Yang, Liuzhen Su, Peng Zhai, Chixiao Chen, Lihua Zhang
Backpropagation through Time and Space: Learning Numerical Methods with Multi-Agent Reinforcement Learning
Elliot Way, Dheeraj S. K. Kapilavai, Yiwei Fu, Lei Yu
Is it all a cluster game? -- Exploring Out-of-Distribution Detection based on Clustering in the Embedding Space
Poulami Sinhamahapatra, Rajat Koner, Karsten Roscher, Stephan Günnemann