telepathboy/sentinel_nexus repository overview

⁠Efficient Edge AI for Security (CODE NAME: Sentinel Nexus)

⁠Project Overview

This is a Fayetteville State University Intelligent Systems Laboratory (ISL) project called Efficient Edge AI for Security that incorporates a Hybrid-AI-Model known as Sentinel Nexus to detect behavioral changes within a crowd.

⁠Project Goal

To create a prototype Hybrid-AI-Model that can be used within edge devices to achieve the following goals within near-real-time calculations:

• Identify Objects
• Track Objects
• Identify Object Velocity
• Interpolate 3D Objects From 2D Imagery
• Predict Crowd Behaviours

⁠Docker Commands

⁠Pulling the Repository

Ensure you choose the correct tag for your platform (amd64 or arm64v8)

docker pull telepathboy/sentinel_nexus:<tag>

⁠Running the Docker Image

This will remove the docker container after exit.

docker run --rm -it telepathboy/sentinel_nexus:<tag>

⁠Running The Models

⁠DeepSORT

Single File Command

This will run the model on a single file within the ./Videos/ folder

python3 run_full_deep_sort.py --video=./Videos/<file_name> --model=resources/networks/mars-small128.pb --min_confidence=0.3 --nn_budget=100 --display=False --save_video=True

Multi Video Command

This will run the model on all video files within ./Videos/ folder

python3 run_full_deep_sort.py --video_folder=./Videos/ --model=resources/networks/mars-small128.pb --min_confidence=0.3 --nn_budget=100 --display=False --save_video=True