Sistema de reconhecimento facial com base em técnicas de aprendizado de máquina
The work proposes studies in the field of machine learning for facial recognition. The implementation of a personal identification system is sought to compare information flows contained in datasets obtained from the processing of videos and photos of individuals as a mechanism for fraud prevention....
| Autor principal: | SANTOS, Vagner Vieira dos |
|---|---|
| Grau: | Artigo |
| Idioma: | por |
| Publicado em: |
2022
|
| Assuntos: | |
| Acesso em linha: |
https://bdm.ufpa.br:8443/jspui/handle/prefix/3983 |
| Resumo: |
|---|
|
The work proposes studies in the field of machine learning for facial recognition. The implementation of a personal identification system is sought to compare information flows contained in datasets obtained from the processing of videos and photos of individuals as a mechanism for fraud prevention. Neural Networks technology is used to create embeddings (vector representation of facial features) of faces by extracting faces from photos. Thus, models previously trained in massive databases, such as OpenFace and Facenet, are used, as they are validated and demonstrate scientific results with a high rate of effectiveness. It is worth mentioning that the procedures were applied and compared with different classifiers such as Support Vector Machine (SVM), k-Nearest Neighbors (KNN) and Convolutional Neural Network (ConvNet). The acquired results were based on the comparison process between Python programming via Jupyter Notebook and Orange Software version 3.29.3. The tests carried out showed positive and promising results. |