YOLACT_training

Use YOLACT training instance segmentation, the training set data is downloaded from roboflow Raccoon Dataset and then labeled.

The labeled data is placed in the raccoon_dataset folder.

Create dataset

Before training, first convert the training data into COCO format.

Execute create_dataset.ipynb to convert the training data format.

Execute training.ipynb for training.

The trained model is placed in weights

The predicted results are as follows

Press F12 on the Colab page, paste the following Code in the Console and press Enter

function ConnectButton(){
console.log("Connect pushed"); 
document.querySelector("#connect").click() 
}
setInterval(ConnectButton,60000);

The detailed process can refer to my Medium: YOLACT 訓練教學