This is an experimentation with evolutionary computing applied to neural networks using TensorFlow.js
npm install tfjs-evolutionimport { DisplayPanelComponent } from 'tfjs-evolution';
imports: [DisplayPanelComponent]
@ViewChild(DisplayPanelComponent) child!:DisplayPanelComponent;
typescript
classes_names=[
"Canário-da-Terra",
"Pardal",
"Bem-Te-Vi",
"Suiriri-Cavaleiro",
"Lavadeira-Mascarada",
"Anu-Branco",
"Beija-Flor-Tesoura",
"Casaca-de-Couro-da-Lama",
"Garibaldi",
"Periquitão",
"Pombo-Doméstico",
"Quero-Quero",
"Rolinha-Roxa",
"Sabiá-Do-Campo",
"Urubu-de-Cabeça-Preta"
]
``
Make the call for loading images. You could most likely make the call at the constructor.
`typescript
ngAfterViewInit(): void {
const options={
base: "./assets/dataset",//base where are your iamges
file_name:"image",//how your images are named
file_extension:"jpeg" //extension used
};
//Finally, call the load images methods
this.child.loadImages(5, options)
}
`
Finally, do not forget to add the HTML code
`html
Images used on the training
``