train_detectron2_instance_segmentation
About
Train Detectron2 instance segmentation models
Train your custom Detectron2 instance segmentation models.
🚀 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 workflowwf = Workflow()# Add dataset loadercoco = wf.add_task(name="dataset_coco")coco.set_parameters({"json_file": "path/to/annotation/file.json","image_folder": "path/to/image/folder","task": "instance_segmentation",})# Add training algorithmtrain = wf.add_task(name="train_detectron2_instance_segmentation", auto_connect=True)# Launch your training on your datawf.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
-
model_name (str) - default 'COCO-InstanceSegmentation/mask_rcnn_R_50_FPN_3x': Name of the pre-trained model. Additional model available:
- COCO-InstanceSegmentation\mask_rcnn_R_101_C4_3x
- COCO-InstanceSegmentation\mask_rcnn_R_101_DC5_3x
- COCO-InstanceSegmentation\mask_rcnn_R_101_FPN_3x
- COCO-InstanceSegmentation\mask_rcnn_R_50_C4_1x
- COCO-InstanceSegmentation\mask_rcnn_R_50_C4_3x
- COCO-InstanceSegmentation\mask_rcnn_R_50_DC5_1x
- COCO-InstanceSegmentation\mask_rcnn_R_50_DC5_3x
- COCO-InstanceSegmentation\mask_rcnn_R_50_FPN_1x
- COCO-InstanceSegmentation\mask_rcnn_R_50_FPN_1x_giou
- COCO-InstanceSegmentation\mask_rcnn_R_50_FPN_3x
- COCO-InstanceSegmentation\mask_rcnn_X_101_32x8d_FPN_3x
- LVISv0.5-InstanceSegmentation\mask_rcnn_R_101_FPN_1x
- LVISv0.5-InstanceSegmentation\mask_rcnn_R_50_FPN_1x
- LVISv0.5-InstanceSegmentation\mask_rcnn_X_101_32x8d_FPN_1x
- LVISv1-InstanceSegmentation\mask_rcnn_R_101_FPN_1x
- LVISv1-InstanceSegmentation\mask_rcnn_R_50_FPN_1x
- LVISv1-InstanceSegmentation\mask_rcnn_X_101_32x8d_FPN_1x
-
max_iter (int) - default '100': Maximum number of iterations.
-
batch_size (int) - default '2': Number of samples processed before the model is updated.
-
input_size (int) - default '400': Size of the input image.
-
output_folder (str, optional): path to where the model will be saved.
-
learning_rate (float) - default '0.0025': Step size at which the model's parameters are updated during training.
-
eval_period (int) - default '50': Interval between evalutions.
-
dataset_split_ratio (float) – default '0.8' ]0, 1[: Divide the dataset into train and evaluation sets.
-
config_file(str, optional): Path to config file.
Parameters should be in strings format when added to the dictionary.
from ikomia.dataprocess.workflow import Workflow# Init your workflowwf = Workflow()# Add dataset loadercoco = wf.add_task(name="dataset_coco")coco.set_parameters({"json_file": "path/to/annotation/file.json","image_folder": "path/to/image/folder","task": "instance_segmentation",})# Add training algorithmtrain = wf.add_task(name="train_detectron2_instance_segmentation", auto_connect=True)train.set_parameters({"model_name": "COCO-InstanceSegmentation/mask_rcnn_R_50_C4_1x","batch_size": "2","input_size": "400","learning_rate": "0.0025","dataset_split_ratio": "0.8",})# Launch your training on your datawf.run()
Developer
Ikomia
License
Apache License 2.0
A permissive license whose main conditions require preservation of copyright and license notices. Contributors provide an express grant of patent rights. Licensed works, modifications, and larger works may be distributed under different terms and without source code.
Permissions | Conditions | Limitations |
---|---|---|
Commercial use | License and copyright notice | Trademark use |
Modification | State changes | Liability |
Distribution | Warranty | |
Patent use | ||
Private use |
This is not legal advice: this description is for informational purposes only and does not constitute the license itself. Provided by choosealicense.com.