Products & Publications

Products and Publications

An important goal of ML4Earth is to build and maintain an international community within the AI4EO domain. We are pursuing this goal together with the Space Agency of the German Aerospace Center. We are doing so by providing benchmark data sets as a service to the community, while offering the community opportunities for training and networking.

Benchmark Datasets

We are offering benchmark datasets for a wide range of application scenarios. Benchmark data products are pre-labeled EO datasets that are delivered together with a baseline solution, i.e., pre-trained AI models. With such benchmarks, the user does not need to start from scratch and train their own AI models, which can become very time-consuming and resource-intensive. Instead, developers can use the benchmarks as a head start and directly apply the delivered models to their science domain or advance the training of their own AI models.

Click Here to Visit EARTHNETS

Community Building

Every year, we are organizing workshops, each covering one of the research domains, and hackathons to foster knowledge exchange democratize AI methods for EO applications. With these events, we share our methodological expertise to enable more researchers to exploit Copernicus and other EO data sources.

The DLR Space Agency is hosting a Slack Channel to build the ML4Earth Community. At this moment, more than 170 students, research staff, and stakeholders have joined the “ML4Earth” channel to discuss, ask questions, network, or have informal exchange.

Would you like to join the channel?

Just send a short e-mail to Dr. Matthias Kahl at matthias.kahl@tum.de

ML4Earth Community Building Symbol Image

Publications

Publications of our research team will be listed here.
The visualization of five data modalities in MDAS. From left to right: the Sentinel-1 image (VH band in dB), the Sentinel-2 image (RGB: 665, 560, 490 nm), the OSM land use layer, the HySpex image (RGB: 691, 569, 445 nm), and the DSM.
MDAS: a new multimodal benchmark dataset for remote sensing (2023)
January 9, 2023
Jingliang Hu, Rong Liu, Danfeng Hong, Andrés Camero, Jing Yao, Mathias Schneider, Franz Kurz, Karl Segl, and Xiao Xiang Zhu