Tutorial on Embedding Projector with our own feature vector.

Tensorflow is a one of the most popular free and open source machine learning library which helps you to do all kind of machine learning and deep learning projects. It can be used for various tasks including training, inferencing, visualising ML and DL models. It also comes up with wide variety of pretrained models.

Now TensorBoard is the visualization toolkit for tensorflow. It helps us for the visualization of various machine learning experiements.

What can we do with TensorBoard?

  • TB helps us to tracks various metrics such as loss and accuracy for ML training.
  • Model Graphs can be visualized.
  • We…


How to configure GPU with Tensorflow in Ubuntu 18.04. What version of CUDA and CUDAnn to choose?

I am planning to make this story as simple and easy. You are probably reading this because you wasted some time in configuring GPU for you laptop or PC to work with tensorflow.
I will try to cover more generally. But the implementation and the configuration that I am focusing on is with Ubuntu-18.04.

Source

First of all what all things we need?

  1. We need a system with GPU. Of course you have that, that is why you are here.
  2. We need Nvidia GPU drivers. For the latest use cases we require 450.0 + versions.
  3. We need to install CUDA toolkit.


How to get dataset for masked face recognition

This short story will give you a walkthrough on how to do masked face recognition. I tried various methods for doing masked face recongition but nothing worked well. Even this methodology is also not giving us performance similar to an ordinary masked face recognition but still better than others.

Photo by Pille-Riin Priske on Unsplash

With the pandemic out break, AI powered smart survellience camera face struggling times to detect the face of people who always wear mask. So there is a necessity to track people no matter which mask they use.

How can we solve this? The only way is to do just as the…


This is a short story on how to install face_recognition on Ubuntu without any errors.

Photo by XPS on Unsplash

I was facing some difficulty in installing face-recognition module. The solution was simple. Go the original repos and see how it works. I hope this will work for you too.

I installed everything on an Ubuntu docker container. So I had to configure many more stuffs which I am not sharing. Don’t worry. This includes pretty basic things like installing python3, pip, dev packages even git too.

For installing face-recognition module for Ubuntu 18.04: (Try for other OS, I used this for 18.04 and it worked)

  1. Install cmake:

pip install cmake

After cmake is successfully installed

2. Optional: Install git…


How society, demographics and technology are changing the world of FM.

The accounted global facility management market size in 2019 was 1.24 trillion USD and is predicted to reach 1.62 trillion USD by 2027. This accounts for a 4.0% increase in the Compound Annual Growth Rate (CAGR).

Companies outsourcing Facilities Management

Source: Mckinsey

Several economic factors have made outsourcing more relevant for facilities management. Facilities management is ripe for disruption: it lags behind other functions such as production equipment maintenance by both digital maturity and penetration of technology. Although technology is available for facilities management, several obstacles have inhibited adoption, such as a lack of digital skills within the function, other priorities for leadership, and a focus…


An Introduction to the OAK-D and OpenCV AI Competition

The gadget for the growing fields of computer vision and 3D perception which packs a lot of capabilities into a small unit.

I started to explore OpenCV AI Kit D (OAK-D) because of the OpenCV AI Competition. This is an international Artificial Intelligence competition celebrating OpenCV’s 20th anniversary. A total prize worth of $400K will be awarded with global and regional winners. This international competition is sponsored by Azure and Intel. In this challenge teams has to work on real life problems using OAK-D.

What is OAK-D?

OAK is a modular, open-source ecosystem composed of MIT-licensed hardware, software, and AI training — that…


Cheat Sheet for OpenCV — Part 4, Finding and Drawing Contours of an image.

Photo by Steinar Engeland on Unsplash

We are going to see how to identify contours in openCV. Contours are the line or points that join the continuous points of the boundary of an object. Contours are very essential tools when you get into object detection, shape analysis, object recognition etc.

Let’s start by importing an image with openCV and also changing it to gray scale. We already discussed how to convert an image to gray scale in the previous stories.


Cheat Sheet for OpenCV — Part 3, Translation | Rotation | Resizing | Flipping | Cropping

Photo by Aaron Burden on Unsplash

1. Translation

Translation is the process of shifting an image along the x and y axis. These translation can be used as data augmentation steps if you are working with deep learning projects.

In order to do this we are going to create a translation function , In this function we are going to take inputs of image, x and y to translate. We need to create a translation matrix to create translation. We do this by the help of numpy . We do the translation with the help of cv.warpAffine() function provided by openCV. …


Cheat Sheet for OpenCV — Part 2, Converting to Grayscale | Resizing | Edge Cascading | Blurring | Dialation | Erosion | Cropping

Photo by Nick Femerling on Unsplash

Probably you will be directed to this story from the part-1 of OpenCV. If not I would recommend reading it will give you a better idea. But it is not necessary though.

These are some of the very basic function in openCV which you have to use on whatever computer vision project you are doing.

  1. Converting an image to grayscale image

For converting an RGB image to gray scale image we should use a function called cv.cvtColor() with the argument <img, color_feature(for converting to grayscale it is cv.COLOR_BGR2GRAY)>


Getting started with latest OpenCV. Part by part explaining all the essentials.

OpenCV — Source

Installing Necessary Packages

First you need to install python. It is easy. Since I am using an Ubuntu OS I will be going through it. If you have other OS just search. Google will help you with anything.

sudo apt update
sudo apt -y upgrade
sudo apt install python3.8
sudo apt install -y python3-pip
pip3 install <package_name> ## To install packages.
## Some essentials for the python development
sudo apt install -y build-essential libssl-dev libffi-dev python3-dev

If you want setup a virtual environment so that whatever you install can be secured from other projects. …

Raoof Naushad

More wonderful stories | More wonderful experiences

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store