The importance of non-verbal behaviour and what is conveyed through body language, plays a very important role in our environment. As one of the areas this work focuses on the recognition of emotions for children from 2 to 4 years, because it is important to review the state of emotion that represents a child depending on the environment in which he is immersed. The objective of this article is to evaluate the efficiency of three emotion recognition tools, which are Face++, Microsoft Azure Emotion API and Google Vision API when inferring child attributes. In order to carry out this research, a comparative analysis of the tools used was made with the help of a child psychology specialist. A group of 20 children from the Foundation "Los Chavitos" of MIES in Ecuador were taken as a sample. The experimental results showed that Face++ was more accurate than Microsoft Azure Emotion API and Google Vision API. It is hoped that the data set presented in the results can help pave the way for future research into the use of emotion recognition tools.
|Translated title of the contribution||Comparative analysis to select an emotion recognition tool applying the AHP model|
|Original language||Spanish (Ecuador)|
|Journal||UNIANDES - Episteme|
|State||Published - 13 Jun 2019|