Tools
Tools
Kelly, Lauren (2023) A Comparative Analysis of Emotion Detection Accuracy in AWS Rekognition and Luxand FaceSDK: Balancing Performance and Cost. Undergraduate thesis, Dublin, National College of Ireland.
Preview |
PDF (Bachelor of Science)
Download (1MB) | Preview |
Abstract
In this final year project, a comparison is made with the performance of two cloud-based vision applications, Amazon Web Services (AWS) Rekognition and Luxand FaceSDK, in detecting emotions from a dataset of labelled images. The dataset was compiled by amalgamating three publicly available datasets, providing a diverse and representative sample for testing the emotion detection capabilities of these APIs.
The study aimed to analyse the accuracy of these two platforms, which possess different cost structures for the users, to help stakeholders, make informed decisions when choosing a facial recognition API for their applications.
To conduct this comparison, custom Python code was developed to iterate over the images, call the respective cloud services, and save the emotion detection results to JSON files. The accuracy of the APIs was determined by comparing their outputs to the labels provided within the original dataset.
Face recognition software must be accurate because it has wide-ranging effects on a variety of industries, including security, marketing, and entertainment. To accurately identify people and grant or deny access, security systems, for instance, rely on facial recognition software. Marketing campaigns can be more focused and successful if they consider the emotional reactions of consumers to advertisements. Emotion detection can be used in interactive experiences like virtual reality or gaming in the entertainment industry to increase user engagement.
In conclusion, this project provides a comprehensive comparison of the AWS Rekognition and Luxand FaceSDK platforms in terms of emotion detection accuracy. The results will enable potential users to make informed decisions based on their specific requirements and budget constraints while emphasizing the significance of accuracy in facial recognition applications across various domains.
| Item Type: | Thesis (Undergraduate) |
|---|---|
| Supervisors: | Name Email Moldovan, Arghir-Nicolae UNSPECIFIED |
| Subjects: | Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science B Philosophy. Psychology. Religion > Psychology > Emotions Q Science > QA Mathematics > Electronic computers. Computer science > Computer Systems > Computers > Electronic data processing > Programming languages T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science > Computer Systems > Computers > Electronic data processing > Programming languages |
| Divisions: | School of Computing > Bachelor of Science (Honours) in Computing |
| Depositing User: | Tamara Malone |
| Date Deposited: | 16 Jan 2024 16:41 |
| Last Modified: | 16 Jan 2024 16:41 |
| URI: | https://norma.ncirl.ie/id/eprint/6924 |
Actions (login required)
 |
View Item |