LAURIE VERZEAUX*, JOSSELIN BREUGNOT*, HÉLÈNE MUCHICO, ELODIE AYMARD, BRIGITTE CLOSS
*Corresponding authors
SILAB, Brive, France

Abstract

Artificial intelligence (AI) is a set of theories and techniques capable of simulating human intelligence. Among AI, there are different subsets allowing to automatically classify and quantify a high amount of data. Deep learning is the most recent and can define for itself features allowing to identify and quantify information directly from an image. Two examples of deep learning applications on image analysis in the field of dermocosmetic research will be detailed in this publication.

INTRODUCTION
The artificial intelligence (AI) term was coined by John McCarthy in 1956, during a conference at Darmouth College. For many years, AI models were developed but this discipline took on its full dimension with the emergence of computer science. The proportion of papers in the Scopus database that mention AI or AI-related keywords in the title or abstract now stands at 8%, up from 2% a decade ago, according to an analysis by Nature (1). Today, many applications based on these models, such as search engines, customer service chatbots, social media recommendation algorithms, credit scoring etc. are used on a daily basis.

Among AI, machine learning and deep learning are different subsets (Figure 1).

Machine learning appears in the 1980’s and uses algorithms, such as decision trees or linear regression for example, to represent how different input can be used to predict an outcome. For this approach, the features which will allow automatic classification must be selected beforehand from the data by an operator. This step can be complex and requires knowledge of the subject o ...