Dear Students,
This message is addressed to the students who have expressed interest in
the "Open-Source AI Framework for Thermal Satellite Payload Data
Analysis" proposed project. We thank you for reaching out and we are
writing this message to clarify some points regarding the project and
help you better focus your proposals.
First, please note that the main repository reflects the work of the
research team, while the "/TIRAuxCloud" repository contains the code
directly related to this specific project. At the moment, the available
dataset and modeling code primarily covers the task of cloud detection
using thermal infrared data. In this direction, we are interested in
proposals that focus on:
*
Extensions related to uncertainty quantification and explainability
using state-of-the-art methods e.g., single-pass uncertainty,
Grad-CAM++, DeepSHAP.
*
Extensions beyond traditional supervised methodologies (e.g., change
detection, self-supervised, pretrained large models) which could
enable improved discovery of patterns, anomalies, temporal variations.
Other directions may include completion and expansion of the dataset,
enabling training for tasks beyond cloud detection, together with the
development of a robust preprocessing pipeline for thermal satellite
data to produce model-training-ready datasets.
All submitted proposals will be evaluated based on their technical
clarity, completeness, feasibility, and quality of implementation. The
selected projects will be those that present both strong research value
and realistic implementation plans.
Best regards,
The mentors team
----
Λαμβάνετε αυτό το μήνυμα απο την λίστα: Λίστα αλληλογραφίας και συζητήσεων που απευθύνεται σε φοιτητές developers \& mentors έργων του Google Summer of Code - A discussion list for student developers and mentors of Google Summer of Code projects.,
https://lists.ellak.gr/gsoc-developers/listinfo.html
Μπορείτε να απεγγραφείτε από τη λίστα στέλνοντας κενό μήνυμα ηλ. ταχυδρομείου στη διεύθυνση <gsoc-developers+unsubscribe [ at ] ellak [ dot ] gr>.