Face Recognition and OpenCV

What is face recognition and how does it work? First of all what is recognition? Let's take an example, the first thing that comes to mind when you see an orange fruit is that it is an orange fruit. In simple terms, your mind telling you that this is an orange fruit, is recognition. So, what exactly is face recognition? I'm confident you guessed correctly. You identify your friend Bob when you see him strolling down the street or when you see a photo of him. But, more importantly, how does face recognition work? It's easy to use and understand. Consider this scenario: you meet someone for the first time in your life and you don't recognize him, right? You examine his face, eyes, nose, lips, color, and overall appearance while he speaks or shakes your hand. This is your mind learning or training to recognize that person's face by collecting face data.  The next question is how to implement face recognition using OpenCV and other open source libraries? So, there you have it. You would argue that human minds can easily do these things, but programming them into a machine is tough. Don't worry, it's not the case. Face recognition is as simple as it appears thanks to different open source libraries . Face recognition follows the same coding techniques as the real-life example we examined earlier.

Libraries for Image Pre-processing

Following are Image Pre Processing Libraries. By annotating and labelling photos, these libraries aid in data preparation. Additionally, photos are processed in order to run a machine learning algorithm.

Libraries for Face recognition models

There are various libraries for implementing face recognition models, important and easy to use are listed here.

Facenet

Facenet an important model for face recognition. To learn more click here!

Facial-Emotion-Recognition using Images

Images of faces with happy/sad emotion are provided in the data. Using PCA and LDA we can create a face emotion recognizing model based on dimensionality reduction technique. Check the source code here: Visit to my repo!

Human Activity Recognition

Dataset link: Dataset! Classifying activity(WALKING, WALKINGUPSTAIRS, WALKINGDOWNSTAIRS, SITTING, STANDING, LAYING) of person based on smartphone sensor data using various classical machine learning models like SVM, KNN, Logistic regression and comparing there results. Check the code on: Visit to my repo! Following are libraries which are useful to implement the HAR model and doing preprocessing tasks.

Unsupervised Sentiment Analysis

Implemented sentiment analysis model for movie reviews by extraction of TF-IDF features and Gaussian Mixture Models (GMMs) in python. Dataset: Corpus with 1000 movie reviews. Implementing GMMs and TF-IDF with the help of open source libraries we can develop a unsupervised Sentiment Analysis Model. Check the code on: Visit to my repo! Following are open source libraries which are useful to implement this model.