FRAMEWORK

Innovative software infrastructure that allows the evaluation of the image quality for deep learning segmentation and classification algorithms.

The Framework project has the objective to design and develop a new software infrastructure which allows the evaluation of images quality for deep learning segmentation and classification algorithms, allowing the quality measurement of the images according to the prediction accuracy of different deep learning algorithms.

The innovations promoted by the project aims at the evaluation of the image quality from the satellites after being compressed and restored, to analyse to what extent the images can be modified so that they do not suffer a significant deterioration.

Framework, a project coordinated by Satellogic, counts with the participation of the Multimedia Technologies Unit of Eurecat which works in tasks focused to the design and implementation of Deep Learning-based Computer Vision algorithms, together with the Group on Interactive Coding of Images (GICI) of the Autonomous University of Barcelona (UAB).