skilobyte/distributed-face-detection

A distributed system that detects face in a video and draws Delaunay triangles over it.

Distributed Face Detection

A distributed system that detects faces in a video and draws Delaunay triangles over it.

Features

• Easy Setup - Seamless installation process through Docker images.
• Highly Scalable - Easily scale the number of worker nodes to match your desired workload with minimal configuration changes.
• Handles Multiple Tasks - Capable of efficiently handling multiple video requests simultaneously.
• Robust Logging - Incorporates comprehensive logging functionalities to track and monitor activities effectively.

Setup

Requirements

• Python
• IDE of your choice
• Docker
• PostgreSQL

Instructions

1. Pull the repo and open the terminal in the project directory

2. Change directory to Kafka folder cd kafka

3. Run the docker compose file with the command docker compose up -d

4. Run the file initialKafkaConfig.py with the command python initialKafkaConfig.py

5. Change the directory to workerNode folder using the command cd ../workerNode

6. Change the IP address in the .env file to the IP address of the machine running Kafka.

7. Build a Docker image with the command: docker build -t node:1.0 .

8. Change the machine ID as required in the .env file and build another docker image with the command: docker build -t node:1.1 .

9. Create a container using the command docker run --name worker1 -d -p 9090:9093 node:1.0 and docker run --name worker2 -d -p 9099:9093 node:1.1 Now the message queue (Kafka) and two workers are ready.

10. Change to the machineNode folder using the command cd ../machineNode, and change the IP address in .env to the IP address of the machine running Kafka.

11. Run the server.py to host the website in localhost:5000, upload the video and click send.

12. The resulting video will be displayed after a few seconds with complete statistics.

License