Skip to content

HackAI 2025 | ISS Docking Port Detection and Distance Regression AI

License

Notifications You must be signed in to change notification settings

RandomKiddo/HackAI2025

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

51 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

HackAI 2025 | ISS AI Docking Locator and Distance Regressor

GitHub License GitHub repo size

THIS REPO MAY HAVE BEEN CHANGED OR UPDATED SINCE THE ORIGINAL HACK AI VERSION
SEE THE HACK AI RELEASE FOR THAT VERSION


Team

LeTeam: Neil Ghugare, Pranav Moola, Nishanth Kunchala, and Jacob Balek


Acknowledgements

The data was collected on Kaggle, taken from a an AICrowd challenge dataset, located here.

We used public-use STL files of the ISS and Dragon shuttle, located here and here, respectively.


HackAI 2025 Model

The HackAI 2025 model was created as a part of the 25-hour hackathon. We created a MHN regressing three values (distance, x-coordinate of the docking port, and y-coordinate of the docking port) through three separate heads. The backbone of the model was the MobileNetV3Small model provided by Keras.

Loss and Val Loss of the MHN HackAI Model

Top Layer of the MHN (MobileNet not included)

Alongside that, a 3D visualization of the model outputs was created using STL files of the ISS and SpaceX Dragon capsule (provided).

Test Gif of HackAI Model


Back to Top

This page was last edited on 02.23.2025