Intro

Hot Signals is a rating system based on how urgent & maintenance is needed on provided imagery of a cell tower.

It was first created from Verizon's WeHack25 Prompt on cell tower detection.

How It's Ran

Ran locally on my desktop using flask.
Can upload files through the website.
Files are then put through 4 different test.
    Rust: Find corrosion on cell tower
    Buildings: Find buildings in background
    Unknown Objects: Find odd or unusual objects on the cell tower
    Antennas: Find antenna and placement
Using this data, a rating based on maintenance can be calculated. The higher the rating the higher the maintenance.
On the website, the end results are presented.

Demo

Credits

Rust Corrosion Detection: 
@misc{
    rush_corrosion_detection_dataset,
    title = { rush_corrosion_detection Dataset },
    type = { Open Source Dataset },
    author = { yolocorrosionrustdetection },
    howpublished = { \url{ https://universe.roboflow.com/yolocorrosionrustdetection/rush_corrosion_detection } },
    url = { https://universe.roboflow.com/yolocorrosionrustdetection/rush_corrosion_detection },
    journal = { Roboflow Universe },
    publisher = { Roboflow },
    year = { 2024 },
    month = { sep },
    note = { visited on 2025-04-23 },
}

All Libraries Used

(Or What I Needed to Install / For Reference for Me Later)

Virtual Enviroment (venv):
    python3 -m venv <nameofvenv>
    source venv/bin/activate
    deactivate (when done)
Flask              : pip install flask
Gemini             : pip install google-gemini
Ultralytics/OpenCV : pip install ultralytics 
OpenCV             : pip install opencv-python
Inference          : pip install inference (maybe)