RoboflowInstanceSegmentationModel¶

Run inference on an instance segmentation model hosted on or uploaded to Roboflow.

You can query any model that is private to your account, or any public model available on Roboflow Universe.

You will need to set your Roboflow API key in your Inference environment to use this block. To learn more about setting your Roboflow API key, refer to the Inference documentation.

Properties¶

Name	Type	Description	Refs
`name`	`str`	Unique name of step in workflows.	❌
`model_id`	`str`	Roboflow model identifier.	✅
`class_agnostic_nms`	`bool`	Value to decide if NMS is to be used in class-agnostic mode..	✅
`class_filter`	`List[str]`	List of classes to retrieve from predictions (to define subset of those which was used while model training).	✅
`confidence`	`float`	Confidence threshold for predictions.	✅
`iou_threshold`	`float`	Parameter of NMS, to decide on minimum box intersection over union to merge boxes.	✅
`max_detections`	`int`	Maximum number of detections to return.	✅
`max_candidates`	`int`	Maximum number of candidates as NMS input to be taken into account..	✅
`mask_decode_mode`	`str`	Parameter of mask decoding in prediction post-processing..	✅
`tradeoff_factor`	`float`	Post-processing parameter to dictate tradeoff between fast and accurate.	✅
`disable_active_learning`	`bool`	Parameter to decide if Active Learning data sampling is disabled for the model.	✅
`active_learning_target_dataset`	`str`	Target dataset for Active Learning data sampling - see Roboflow Active Learning docs for more information.	✅

The Refs column marks possibility to parametrise the property with dynamic values available in workflow runtime. See Bindings for more info.

Available Connections¶

Check what blocks you can connect to RoboflowInstanceSegmentationModel.

inputs: AbsoluteStaticCrop, Crop, RelativeStaticCrop
outputs: ActiveLearningDataCollector, Crop, DetectionOffset, DetectionsConsensus, DetectionFilter

The available connections depend on its binding kinds. Check what binding kinds RoboflowInstanceSegmentationModel has.

Bindings

input
- images (Batch[image]): Reference at image to be used as input for step processing.
- model_id (roboflow_model_id): Roboflow model identifier.
- class_agnostic_nms (boolean): Value to decide if NMS is to be used in class-agnostic mode..
- class_filter (list_of_values): List of classes to retrieve from predictions (to define subset of those which was used while model training).
- confidence (float_zero_to_one): Confidence threshold for predictions.
- iou_threshold (float_zero_to_one): Parameter of NMS, to decide on minimum box intersection over union to merge boxes.
- max_detections (integer): Maximum number of detections to return.
- max_candidates (integer): Maximum number of candidates as NMS input to be taken into account..
- mask_decode_mode (string): Parameter of mask decoding in prediction post-processing..
- tradeoff_factor (float_zero_to_one): Post-processing parameter to dictate tradeoff between fast and accurate.
- disable_active_learning (boolean): Parameter to decide if Active Learning data sampling is disabled for the model.
- active_learning_target_dataset (roboflow_project): Target dataset for Active Learning data sampling - see Roboflow Active Learning docs for more information.
output
- prediction_type (Batch[prediction_type]): String value with type of prediction.
- predictions (Batch[instance_segmentation_prediction]): 'predictions' key from Roboflow instance segmentation model output.
- parent_id (Batch[parent_id]): Identifier of parent for step output.
- image (Batch[image_metadata]): Dictionary with image metadata required by supervision.

Example JSON definition of RoboflowInstanceSegmentationModel step

{
    "name": "<your_step_name_here>",
    "type": "RoboflowInstanceSegmentationModel",
    "images": "$inputs.image",
    "model_id": "my_project/3",
    "class_agnostic_nms": true,
    "class_filter": [
        "a",
        "b",
        "c"
    ],
    "confidence": 0.3,
    "iou_threshold": 0.4,
    "max_detections": 300,
    "max_candidates": 3000,
    "mask_decode_mode": "accurate",
    "tradeoff_factor": 0.3,
    "disable_active_learning": true,
    "active_learning_target_dataset": "my_project"
}