train_torchvision_faster_rcnn

Train Faster RCNN object detection model.

🚀 Use with Ikomia API

1. Install Ikomia API

We strongly recommend using a virtual environment. If you're not sure where to start, we offer a tutorial here.

pip install ikomia

2. Create your workflow

from ikomia.dataprocess.workflow import Workflow

# Init your workflow
wf = Workflow()    

# Add dataloader
coco = wf.add_task(name="dataset_coco")

coco.set_parameters({
    "json_file": "path/to/json/annotation/file",
    "image_folder": "path/to/image/folder",
    "task": "detection",
}) 

# Add training algorithm
train = wf.add_task(name="train_torchvision_faster_rcnn", auto_connect=True)

# Launch your training on your data
wf.run()

☀️ Use with Ikomia Studio

Ikomia Studio offers a friendly UI with the same features as the API.

• If you haven't started using Ikomia Studio yet, download and install it from this page.

• For additional guidance on getting started with Ikomia Studio, check out this blog post.

📝 Set algorithm parameters

• epochs (int) - default '15': Number of complete passes through the training dataset.
• batch_size (int) - default '8': Number of samples processed before the model is updated.
• learning_rate (float) - default '0.005': Step size at which the model's parameters are updated during training.
• weight_decay (float) - default '0.0005': Amount of weight decay, regularization method.
• momentum (float) - default '0.9: Optimization technique that accelerates convergence.
• input_size (int) - default '224': Size of the input image.
• classes (int) - default '2': Number of classes
• export_pth (bool) - default 'True'
• export_onnx (bool) - default 'False'
• output_folder (str, optional): path to where the model will be saved.
• num_workers (int) - default '0': How many parallel subprocesses you want to activate when you are loading all your data during your training or validation.

Parameters should be in strings format when added to the dictionary.

from ikomia.dataprocess.workflow import Workflow

# Init your workflow
wf = Workflow()    

# Add dataloader
coco = wf.add_task(name="dataset_coco")

coco.set_parameters({
    "json_file": "path/to/json/annotation/file",
    "image_folder": "path/to/image/folder",
    "task": "detection",
}) 

# Add training algorithm
train = wf.add_task(name="train_torchvision_faster_rcnn", auto_connect=True)
train.set_parameters({
    "classes": '2',
    "batch_size": "8",
    "epochs": "5",
    "input_size": "240",
    "momentum": "0.9",
    "learning_rate": "0.005",
    "weight_decay": "0.0005",
    "export_pth": "True",
    "export_onnx": "False",
}) 

# Launch your training on your data
wf.run()