SpaceSense’s AI platform is accessed through a python library enabling data scientists and developers to create geospatial solutions using artificial intelligence faster without specific expertise. This increases time to market, increases productivity of existing teams and reduces manual repetitive tasks.
Get in touchAccess numerous data sources ready to be used, without any need for remote sensing knowledge.
Get access to various types of satellite data including optical, radar, hyperspectral from a wide range of providers, pre-processed with your custom settings, and ready to be used.
Easily integrate a large catalogue of publicly available geospatial datasets like biodiversity datasets, land use classification maps, soil moisture sensors, weather & many more to speed up your analysis.
Merge multiple datasets into multidimensional data cubes in a few minutes through our features
The main benefit of the SpaceSense library is the ability to very quickly input your own data and to combine it with other sources of geospatial data in a seamless and protected way. Our data input features will accept any type of geolocalized data.
The fusion features allows you to combine information from multiple sources (our satellite ARD, our public geospatial datasets or your proprietary geospatial data) in a few minutes and very easily. You have access to a wide range of settings to control your AOIs, your period, your CRS and much more.
When combining satellite sources, you sometimes need to deal with different resolutions and/or projections. Our harmonization module will help you create coherence between them.
Accelerate your machine learning preparation process with our dedicated features for data preparation
To avoid losing countless hours of training with a faulty dataset, this feature contains several tools to identify outliers, missing values, classes and bounds overlaps and other such tools, ensuring your dataset’s quality.
Labelled data in EO is hard to come by. This feature offers several ways to artificially increase the size of your labelled data using various AI & computer vision techniques by a factor 4x or more.
Depending on the of ML/AI task that you need to perform, you’ll have to prepare the data in a specific way. Our data loaders will do that automatically for you, you just have to specify which task you want to do.
This feature enables faster and efficient feature engineering to get the best ML prediction performance by encapsulating earth observation knowledge.
This module supports the previous ones to make sure that they are running smoothly and provide you with the best experience possible
All of our features are run on SpaceSense’s cloud infrastructure to ensure the scalability and speed necessary to process any type of dataset sizes, from houses to continents.
When it comes to storage, the choice is yours. We offer the possibility to store the results of your work on our cloud infrastructure, but you can also decide to download it locally, or to pull it into your own cloud, for further processing.
We designed our service to be at a level of quality allowing you to use our tools in a production environment, with the associated speed and SLAs.