Press "Enter" to skip to content

Author: Johannes S. Fischer

Facial Emotion Recognition with CNNs in TensorFlow

In a past university project my teammates and I were researching which parts of a face are the most discriminative for a convolutional neural network to classify emotions. For that we trained several facial emotion recognition models with TensorFlow on two databases: FER+ and RAF-DB. The final model architecture as…

Comments closed

Class Activation Maps

A method to visualize internal representations of a convolutional neural network and perform object localization without bounding box annotations. For a university project we investigated the question of where a deep convolutional neural network (CNN) looks, when classifying emotions. For this purpose, we evaluated three visualization methods for CNNs, one…

Comments closed

Localizing Tiago Robot with a Particle Filter in Python & ROS

This project was done for the Mobile Robotics course in the Intelligent Interactive Systems master’s program at Pompeu Fabra University, Barcelona. In the following I will give a short overview of how I approached the task of implementing a localization module and present the results. The code for this project…

Comments closed

Intuitive Explanation of the Kullback-Leibler Divergence

In this post I would like to write and give some intuition about the Kullback-Leibler Divergence, which is a measure of how different two probability distributions over the same random variable are. I’ll start by giving an intuitive explanation of the entropy and then derive the Kullback-Leibler Divergence from it.…

Comments closed

Object Detection Part 2: Localization and Classification

Simple object localization and classification using a convolutional neural network build with Tensorflow/Keras in Python. In my previous post I wrote about a simple object localization problem: predicting the bounding box of a single rectangle on neutral background. However, this approach was limited to a single shape and could not…

Comments closed

Correlation-based Feature Selection in Python from Scratch

Including feature selection methods as a preprocessing step in predictive modeling comes with several advantages. It can reduce model complexity, enhance learning efficiency, and can even increase predictive power by reducing noise. In this blog post I want to introduce a simple python implementation of the correlation-based feature selection algorithm…

Comments closed