Deep Learning

A series of articles dedicated to deep learning. All codes and exercises of this section are hosted on GitHub in a dedicated repository :


The Rosenblatt’s Perceptron : An introduction to the basic building block of deep learning.

Multilayer Perceptron (MLP) : The MLP, or Artificial Neural Network, is a widely used algorithm in Deep Learning. What is it ? How do they learn ?

Full introduction to Neural Nets : A full introduction to Neural Nets from the Deep Learning Course in Pytorch by Facebook (Udacity).

How do Neural Networks learn? : Dive into feedforward process and back-propagation.

Activation functions in DL : An overview of the different activation functions in Deep Learning, how to implement them in Python, their advantages and disadvantages.

Prevent Overfitting of Neural Netorks : Your model overfits ? One of these techniques should help ! We’ll cover class imbalance, data augmentation, regularization, early stopping, reducing learning rate…


A guide to Inception Architectures in Keras : Inception is a deep convolutional neural network architecture that was introduced for the first time in 2014. It won the ImageNet Large-Scale Visual Recognition Challenge (ILSVRC14).

Xception and the Depthwise Separable Convolutions : Xception is a deep convolutional neural network architecture that involves Depthwise Separable Convolutions. It was developped by Google researchers.

Create an Auto-Encoder using Keras functional API : Autoencoder is a type a neural network widely used for unsupervised dimension reduction. So, how does it work ? What can it be used for ? And how do we implement it in Python ?


Speaker Verification using Gaussian Mixture Model (GMM-UBM) : Speaker biometrics is a field of Speech processing which focuses on identifying a unique speaker from several audio recorings. This can be useful for access control or suspect identification for example.

Speaker Verification using SVM-based methods : Another method relying on Support Vector Machines for Speaker Verification.


Introduction to Natural Language Processing : What is NLP ? What is it used for ?

Text Preprocessing : Preprocessing in Natural Language Processing (NLP) is the process by which we try to “standardize” the text we want to analyze.

Text Embedding with Bag-Of-Words and TF-IDF : In order to analyze text and run algorithms on it, we need to embed the text. The notion of embedding simply means that we’ll conver the input text into a set of numerical vectors that can be used into algorithms. In this article, we’ll cover BOW and TF-IDF, two simple techniques for embedding.

Text Embedding with Word2Vec : A deeper dive into the state of the art embedding technique : Word2Vec.

Data Augmentation in NLP : Details of the implementation of “Easy Data Augmentation” paper.

I trained a Network to Speak Like Me (and it’s funny) : Over the course of the past months, I wrote over 100 articles on my blog. That’s quite a large amount of content. An idea then came to my mind : train a language generation model to speak like me. Or more specifically, to write like me.

Few-Shot Text Classification with Human in the Loop : This article addresses the task of classifying texts when we have few training examples.

Improved Few Short Text Classification : As an extension of the previous article, I propose a method that leverages both Data Augmentation and better classifiers.

Character-level LSTMs for Gender Classification from First Name : Implementation of the paper “Predicting the gender of Indonesian Names” on names given in France and in the US using bi-directional character-level LSTMs architecture. Achieved 90% accuracy.


Introduction to Computer Vision : What is Computer Vision ? What are the main concepts ? When should it be used ?

Image Formation and Filtering : How are images formed ? Filters can be applied on the image to extract information. What filters can we use ?

Advanced Filtering and Transformations : In this article, we’ll cover advanced filtering and image transformation techniques.

Local features, Detection, Description and Matching : Local features are used for object tracking for example. We’ll see how to implement them, and cover othe topics.

Images Alignment : When you take a panorama, the image needs to be aligned. How is it done ?

Convolutional Neural Networks (CNN) : CNNs changed the field of Computer Vision. How do CNNs work ? What can they be used for ?

A full guide to face detection : Face Detection using Cascade Classifier, Histogram of Oriented Gradients and Convolutional Neural Networks.

How to use OpenPose on macOS ? : OpenPose is a C++ / Python library for Pose Estimation. Let’s see how to use it in macOS !

Emotion Analysis WebApp : We built a Web App using Flask to analyze job seeking candidates emotions.

Like it? Buy me a coffeeLike it? Buy me a coffee