Background
Distinguished for its expertise in artificial intelligence and facial and emotional recognition technology, Golabs was contacted by its client, Dr. Abel Méndez Porras. Dr. Porras, a university professor with a doctorate in computer science and over 7 years of experience working in AI— for a unique project. Aware of Golabs outstanding reputation, Dr. Porras reached out for assistance with an idea that involved facial recognition technology and children. Dr. Porras wanted to use AI to explore the relationship between children’s feelings or moods and their ability to learn and do well in school, but he needed Golabs to help make his project a reality. Golabs provided the technological tools and created a plan to provide the right technology to help make learning a more positive experience for children.
Challenge
Golabs first implemented artificial intelligence technologies and machine learning to train software to read human faces and expressions in order to perceive human emotions from large categorized datasets. The project required large amounts of video for accuracy. The video processing component of the project required a high processing capacity from the servers and maximum optimization of resources to get the best results. This was achieved through distributed architecture processing that permitted us to scale horizontally instead of vertically.
Objective
To create a system capable of analyzing videos in real-time (or pre-recorded), identifying faces and saving information about the emotions of each individually identified person. In this case, the system would be used in a preschool learning environment.
At the preschool age, learning is paramount as children’s brains are developing critical learning and thinking skills. The technology is designed to read children’s emotions, which impact their performance in school and also attest to the type of experience they’re receiving. A better understanding of a child’s reactions in real-time allows for teachers to make adjustments to his/her teaching styles. These shifts can quickly ameliorate a child’s learning experience with lifelong positive effects.
Strategy & Execution
The project consisted of two phases. The first phase was the construction of an MVP or a demo product, which the client could use to make sure that everything was functioning as expected. In this phase, changes could be made quickly making each iteration better than the previous. Golabs used AI tools and AWS architecture, agile development methods and weekly client meetings to work effectively.
The project consisted of two phases. The first phase was the construction of an MVP or a demo product, which the client could use to make sure that everything was functioning as expected. In this phase, changes could be made quickly making each iteration better than the previous. Golabs used AI tools and AWS architecture, agile development methods and weekly client meetings to work effectively.
The second phase was the delivery of a 100% functional web platform and app with cloud hosting. The app requires a user to upload photos and some personal data to accompany the photos that will allow the app to track the individuals. Once this information has been received, the user can then either upload video or record with the device. After ending the video session, a user can watch the video and the software will show a percentage of time for each feeling that an individual experienced during the duration of the video. The software also has an option to display a color related to the feeling and a scale that estimates the intensity of the emotion. The available emotions for detection are anger, contempt, repulsion, fear, happiness, neutral, sadness, and surprise.
Services provided
The project consisted of two phases. The first phase was the construction of an MVP or a demo product, which the client could use to make sure that everything was functioning as expected. In this phase, changes could be made quickly making each iteration better than the previous. Golabs used AI tools and AWS architecture, agile development methods and weekly client meetings to work effectively.
Using cutting-edge facial recognition technologies, Golabs professional AI team created a platform that is capable of detecting a variety of human emotions from videos. In this case, the platform will be used to help understand children’s moods as they are learning in the classroom. The platform was created to leverage emotion to better understand how children react to subjects and teaching styles. The hope of the project is to create a better learning environment and a lifelong love and enjoyment of learning.
- C#
- Asp.net Core
- Entity Framework Core
- Azure cognitive services
- Postgres
- OpenCV
- Hangfire
- Angular 8
