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

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
}