Facial emotion detection is a common issue focused on in the field of cognitive science. An attempt to understand what exactly we as humans see in each other that gives us insight into other emotions is a challenge we can approach from an artificial intelligence side. While I don't have enough experience in psychology or even artificial intelligence to determine these factors, we can always start off by building a model to determine at least the start of this question. Fer2013 is a dataset with pictures of individuals labeled with the emotions of anger, happiness, surprise, disgust, and sadness. When testing humans on the dataset to correctly identify the facial expression of a set of pictures within the set, the accuracy is 65%.

Sentiment Analysis is already widely used by different companies to gauge consumer mood towards their product or brand in the digital world. However, in the offline world, users are also interacting with the brands and products in retail stores, showrooms, etc., and solutions to measure users' reactions automatically under such settings has remained a challenging task. Emotion detection from facial expressions using AI can be a viable alternative to automatically measure consumers' engagement with their content and brands. In this post, we will discuss how such a technology can be used to solve a variety of real-world use-cases effectively. Car manufacturers around the world are increasingly focusing on making cars more personal and safe for us to drive.

Emotion being a subjective thing, leveraging knowledge and science behind labeled data and extracting the components that constitute it, has been a challenging problem in the industry for many years. With the evolution of deep learning in computer vision, emotion recognition has become a widely-tackled research problem. In this work, we propose two independent methods for this very task. The first method uses autoencoders to construct a unique representation of each emotion, while the second method is an 8-layer convolutional neural network (CNN). These methods were trained on the posed-emotion dataset (JAFFE), and to test their robustness, both the models were also tested on 100 random images from the Labeled Faces in the Wild (LFW) dataset, which consists of images that are candid than posed. The results show that with more fine-tuning and depth, our CNN model can outperform the state-of-the-art methods for emotion recognition. We also propose some exciting ideas for expanding the concept of representational autoencoders to improve their performance.