Skip to content

InstanceSegmentationModel

Run an instance segmentation model.

Step parameters

  • type: must be InstanceSegmentationModel (required)
  • name: must be unique within all steps - used as identifier (required)
  • model_id: must be either valid Roboflow model ID or selector to input parameter (required)
  • image: must be a reference to input of type InferenceImage or crops output from steps executing cropping ( Crop, AbsoluteStaticCrop, RelativeStaticCrop) (required)
  • disable_active_learning: optional boolean flag to control Active Learning at request level - can be selector to input parameter
  • confidence: optional float value in range [0, 1] with threshold - can be selector to input parameter
  • class_agnostic_nms: optional boolean flag to control NMS - can be selector to input parameter
  • class_filter: optional list of classes using as filter - can be selector to input parameter
  • iou_threshold: optional float value in range [0, 1] with NMS IoU threshold - can be selector to input parameter. Default: 0.3.
  • max_detections: optional integer parameter of NMS - can be selector to input parameter
  • max_candidates: optional integer parameter of NMS - can be selector to input parameter
  • mask_decode_mode: optional parameter of post-processing - can be selector to input parameter
  • tradeoff_factor: optional parameter of post-processing - can be selector to input parameter
  • active_learning_target_dataset: optional name of target dataset (or reference to InferenceParemeter) dictating that AL should collect data to a different dataset than the one declared by the model

Step outputs:

  • predictions - details of predictions
  • image - size of input image, that predictions coordinates refers to
  • parent_id - identifier of parent image / associated detection that helps to identify predictions with RoI in case of multi-step pipelines
  • prediction_type - denoting instance-segmentation model