Detections Classes Replacement

Class: DetectionsClassesReplacementBlockV1

Source: inference.core.workflows.core_steps.fusion.detections_classes_replacement.v1.DetectionsClassesReplacementBlockV1

Combine results of detection model with classification results performed separately for each and every bounding box.

Bounding boxes without top class predicted by classification model are discarded, for multi-label classification results, most confident label is taken as bounding box class.

Type identifier

Use the following identifier in step "type" field: roboflow_core/detections_classes_replacement@v1to add the block as as step in your workflow.

Properties

Name	Type	Description	Refs
name	str	Enter a unique identifier for this step..	❌
fallback_class_name	str	The class name to be used as a fallback if no class is predicted for a bounding box.	✅
fallback_class_id	int	The class id to be used as a fallback if no class is predicted for a bounding box;if not specified or negative, the class id will be set to 9223372036854775807.	✅

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

Available Connections

Compatible Blocks

Check what blocks you can connect to Detections Classes Replacement in version v1.

• inputs: Keypoint Detection Model, Line Counter, Google Vision OCR, Florence-2 Model, Single-Label Classification Model, Template Matching, Model Monitoring Inference Aggregator, Florence-2 Model, Overlap Filter, CogVLM, OCR Model, Byte Tracker, Detections Transformation, VLM as Detector, Perspective Correction, Line Counter, Detections Stitch, CSV Formatter, OpenAI, Byte Tracker, Clip Comparison, Dynamic Zone, Multi-Label Classification Model, Object Detection Model, Time in Zone, Path Deviation, Slack Notification, SIFT Comparison, Byte Tracker, Detection Offset, Detections Consensus, Detections Stabilizer, Velocity, YOLO-World Model, Gaze Detection, OpenAI, Pixel Color Count, Llama 3.2 Vision, Bounding Rectangle, Detections Filter, Anthropic Claude, Time in Zone, Detections Merge, Moondream2, Segment Anything 2 Model, Webhook Sink, Roboflow Dataset Upload, Roboflow Custom Metadata, Single-Label Classification Model, Path Deviation, Distance Measurement, VLM as Classifier, Local File Sink, Twilio SMS Notification, Stitch OCR Detections, PTZ Tracking (ONVIF).md), Dynamic Crop, Detections Classes Replacement, Object Detection Model, Google Gemini, Email Notification, OpenAI, Image Contours, Instance Segmentation Model, Multi-Label Classification Model, LMM For Classification, VLM as Detector, Instance Segmentation Model, Roboflow Dataset Upload, LMM, SIFT Comparison, Keypoint Detection Model, VLM as Classifier
• outputs: Stability AI Inpainting, Line Counter, Florence-2 Model, Model Monitoring Inference Aggregator, Label Visualization, Florence-2 Model, Corner Visualization, Triangle Visualization, Overlap Filter, Background Color Visualization, Model Comparison Visualization, Byte Tracker, Detections Transformation, Circle Visualization, Perspective Correction, Line Counter, Detections Stitch, Trace Visualization, Byte Tracker, Blur Visualization, Dynamic Zone, Time in Zone, Path Deviation, Byte Tracker, Detections Consensus, Velocity, Detection Offset, Detections Stabilizer, Bounding Rectangle, Detections Filter, Size Measurement, Time in Zone, Roboflow Dataset Upload, Segment Anything 2 Model, Detections Merge, Polygon Visualization, Roboflow Custom Metadata, Mask Visualization, Bounding Box Visualization, Path Deviation, Distance Measurement, Keypoint Visualization, Ellipse Visualization, Crop Visualization, Color Visualization, Pixelate Visualization, Stitch OCR Detections, PTZ Tracking (ONVIF).md), Dynamic Crop, Detections Classes Replacement, Halo Visualization, Dot Visualization, Roboflow Dataset Upload

Input and Output Bindings

The available connections depend on its binding kinds. Check what binding kinds Detections Classes Replacement in version v1 has.

Bindings

• input

  • object_detection_predictions (Union[instance_segmentation_prediction, keypoint_detection_prediction, object_detection_prediction]): The output of a detection model describing the bounding boxes that will have classes replaced..
  • classification_predictions (classification_prediction): The output of classification model for crops taken based on RoIs pointed as the other parameter.
  • fallback_class_name (string): The class name to be used as a fallback if no class is predicted for a bounding box.
  • fallback_class_id (integer): The class id to be used as a fallback if no class is predicted for a bounding box;if not specified or negative, the class id will be set to 9223372036854775807.

• output

  • predictions (Union[object_detection_prediction, instance_segmentation_prediction, keypoint_detection_prediction]): Prediction with detected bounding boxes in form of sv.Detections(...) object if object_detection_prediction or Prediction with detected bounding boxes and segmentation masks in form of sv.Detections(...) object if instance_segmentation_prediction or Prediction with detected bounding boxes and detected keypoints in form of sv.Detections(...) object if keypoint_detection_prediction.

Example JSON definition of step Detections Classes Replacement in version v1

{
    "name": "<your_step_name_here>",
    "type": "roboflow_core/detections_classes_replacement@v1",
    "object_detection_predictions": "$steps.my_object_detection_model.predictions",
    "classification_predictions": "$steps.my_classification_model.predictions",
    "fallback_class_name": "unknown",
    "fallback_class_id": 77
}