isala404 / speculo Goto Github PK

View Code? Open in Web Editor NEW

24.0 5.0 3.0 79.17 MB

Realtime face detection and recognition using deep learning

Home Page: http://speculo.isala.me/

License: MIT License

Python 1.19% Dockerfile 0.02% HTML 0.03% JavaScript 3.58% CSS 0.03% Jupyter Notebook 94.55% TypeScript 0.46% SCSS 0.14%

data-science tensorflow face-recognition reactjs typescript opencv python3 surveillance speculo faces

speculo's Introduction

speculo - Surveillance done smart

Speculo is an open source platform that utilizes a Deep Convolution Inverse Graphics Network for indexing faces found in a video footage, specifically for CCTV surveillance. It allows you to browse through these footages by the faces found in them rather than thousands of frames. The key differentiator between regular face recognition software and Speculo is that Speculo can recognize faces in different angles.

Prerequisites

React
Python 3.8
Docker
NodeJS
OpenCV
MongoDB

Development Setup

Clone the Project

git clone https://github.com/mrsupiri/speculo/tree/master

Add the database configuration

Speculo uses MongoDB as the database therefore create a file called database_config.env in the root folder and add the following environment variables;

- DB_NAME
- DB_USERNAME
- DB_PASSWORD
- DB_HOST

Run it

Use --build only when you run it for the first time or if you have made changes to the code

docker-compose up --build

How to contribute?

If you want to contribute to Speculo, be sure to review the guidelines in CONTRIBUTING.md.

Resources

Here are the links to pages that explain the system in detail.

License

MIT

speculo's People

Contributors

Stargazers

Watchers

Forkers

dinuka-rp v15a1 manojll

speculo's Issues

Backend - Improvement comparator

Improvement
api/comparator

Description
Rewrite REST API logic from Flask to aiohttp to support asynchronous calls

By Akassharjun Shanmugarajah

Frontend - Filter faces based on the total detected time duration.

Allow the user to filter the results of detections by the total time duration that each person has been detected.

The filtering can be done by less than or equal to/ equal to/ greater than or equal to the specified time.

DevOps - Setup a Kubernetes cluster to host the entire project

Frontend - Edit Name & Blacklist status of a person from the Dashboard

Frontend - Implement Live demonstration

Implement Efficient live detection which downscales the image and POSTs to display in canvas

DevOps - Create an automated training pipeline to run continues experiment

Dockerize Postprocessor & Listing Service

Improvement
api/improvement/dockerize

Description
Dockerize the above services.

By Akassharjun Shanmugarajah

DevOps - Export FaceYolo to tensorflow serving

Design - Website User Interface

Design
Website User Interface

Description
Create a user interface for the react website for Speculo.

By Akassharjun Shanmugarajah & Visal Rajapakse

Backend - Authorization for Admin

Feature
api/auth

Description
Modify existing NodeJS RESTful API service for authorizing admin accounts

By Akassharjun Shanmugarajah

Data Science - Export Face Fingerprint to tensorflow servings

Backend - Add faces into the system for tracking

Backend - Gateway

Feature
api/gateway

Description
Write an API gateway to consume all the microservices provided by the system.

By Kushan Shamika

Frontend - Loading status while analyzing footage

Backend - Authorization

Feature
api/authorization

Description
Write a RESTful API service for a user login system with JWT support for authorization.

By Akassharjun Shanmugarajah

Backend - New service to downscale the image using ffmepg

Create a service that takes file object via HTTP and then call preprocessor using a downscaled versions of that video

Implementation Report

DevOps - Create a CD pipeline to package and ship the project to Kubernetes cluster

Data Science - Update the comparator to use the new model

DevOps - Host the proof of concept model in GCP

Frontend - adding multiple faces into the system for tracking

Improve Live Detection

Current demo is relatively slow due to is sending POST requests of full scaled High resolution images. Improve by downscaling the image to a lower resolution before POST for faster processing.

Frontend - Frontend request handling

By Dinuka Piyadigama

DevOps - Create a CI pipeline to build and test the project

Frontend - home page

By Visal Rajapakse & Dinuka Piyadigama

Dependabot couldn't find a Pipfile for this project

Dependabot couldn't find a Pipfile for this project.

Dependabot requires a Pipfile to evaluate your project's current Python dependencies. It had expected to find one at the path: /api/comparator/Pipfile.

If this isn't a Python project, or if it is a library, you may wish to disable updates for it in the .dependabot/config.yml file in this repo.

View the update logs.

Frontend - Dashboard

ui/ dashboard

By Visal Rajapakse & Dinuka Piyadigama

DevOps - Migrate Yolo face to tensorflow 2

Database credentials is exposed in the files

The credentials for the MongoDB is currently written in the respective service files. We have to move them to a .env file.

Frontend - upload video into the system

By Dinuka Piyadigama

Dependabot couldn't find a Pipfile for this project

Dependabot couldn't find a Pipfile for this project.

Dependabot requires a Pipfile to evaluate your project's current Python dependencies. It had expected to find one at the path: /api/postprocessor/Pipfile.

If this isn't a Python project, or if it is a library, you may wish to disable updates for it in the .dependabot/config.yml file in this repo.

View the update logs.

Data Science - Create the proof of concept with face recognition library

Data Science - Create a Face FingerPrinter with keras

Add a README.md and a description for the project

Backend - Preprocessor

Feature
api/preprocessor

Description
Write a preprocessor to open the video sent by the front end frame by frame and process it to return the names and fingerprints of the people present in the video.

By Akassharjun Shanmugarajah