Detections Stabilizer

Class: StabilizeTrackedDetectionsBlockV1

Source: inference.core.workflows.core_steps.transformations.stabilize_detections.v1.StabilizeTrackedDetectionsBlockV1

This block stores last known position for each bounding box If box disappears then this block will bring it back so short gaps are filled with last known box position The block requires detections to be tracked (i.e. each object must have unique tracker_id assigned, which persists between frames) WARNING: this block will produce many short-lived bounding boxes for unstable trackers!

Type identifier

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

Properties

Name	Type	Description	Refs
name	str	Enter a unique identifier for this step..	❌
smoothing_window_size	int	Predicted movement of detection will be smoothed based on historical measurements of velocity, this parameter controls number of historical measurements taken under account when calculating smoothed velocity. Detections will be removed from generating smoothed predictions if they had been missing for longer than this number of frames..	✅
bbox_smoothing_coefficient	float	Bounding box smoothing coefficient applied when given tracker_id is present on current frame. This parameter must be initialized with value between 0 and 1.	✅

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 Stabilizer in version v1.

• inputs: Stability AI Outpainting, Image Slicer, Pixelate Visualization, SIFT Comparison, Image Threshold, Blur Visualization, Image Slicer, Moondream2, Image Blur, Byte Tracker, Camera Calibration, Detections Consensus, Morphological Transformation, Polygon Visualization, Dot Visualization, Time in Zone, PTZ Tracking (ONVIF).md), Background Color Visualization, Distance Measurement, Identify Changes, VLM as Detector, Stability AI Image Generation, Overlap Filter, Corner Visualization, Line Counter, Halo Visualization, Mask Visualization, Detections Merge, Detections Classes Replacement, Trace Visualization, OCR Model, Camera Focus, Color Visualization, Clip Comparison, Identify Outliers, Detections Stitch, Ellipse Visualization, Instance Segmentation Model, Velocity, Model Comparison Visualization, Object Detection Model, Time in Zone, Triangle Visualization, Image Preprocessing, Template Matching, SIFT, Depth Estimation, Detections Filter, Dynamic Crop, Line Counter Visualization, VLM as Detector, Crop Visualization, Image Contours, Grid Visualization, Contrast Equalization, Segment Anything 2 Model, Instance Segmentation Model, Google Vision OCR, Time in Zone, SIFT Comparison, Detections Combine, Classification Label Visualization, Stitch Images, Keypoint Visualization, Absolute Static Crop, Byte Tracker, Perspective Correction, Path Deviation, Bounding Rectangle, Byte Tracker, Pixel Color Count, Polygon Zone Visualization, Dynamic Zone, YOLO-World Model, Image Convert Grayscale, QR Code Generator, Detections Stabilizer, Circle Visualization, Icon Visualization, Object Detection Model, Stability AI Inpainting, Bounding Box Visualization, Detections Transformation, Path Deviation, Detection Offset, EasyOCR, Label Visualization, Reference Path Visualization, Relative Static Crop, Line Counter
• outputs: Roboflow Custom Metadata, Size Measurement, Pixelate Visualization, Blur Visualization, Detections Filter, Dynamic Crop, Crop Visualization, Byte Tracker, Segment Anything 2 Model, Detections Consensus, Time in Zone, Detections Combine, Polygon Visualization, Dot Visualization, Florence-2 Model, Time in Zone, Roboflow Dataset Upload, PTZ Tracking (ONVIF).md), Background Color Visualization, Distance Measurement, Byte Tracker, Perspective Correction, Path Deviation, Roboflow Dataset Upload, Corner Visualization, Overlap Filter, Line Counter, Bounding Rectangle, Dynamic Zone, Halo Visualization, Detections Merge, Mask Visualization, Detections Classes Replacement, Trace Visualization, Stitch OCR Detections, Detections Stabilizer, Circle Visualization, Model Monitoring Inference Aggregator, Icon Visualization, Color Visualization, Stability AI Inpainting, Florence-2 Model, Bounding Box Visualization, Detections Stitch, Ellipse Visualization, Detections Transformation, Path Deviation, Detection Offset, Velocity, Model Comparison Visualization, Label Visualization, Byte Tracker, Time in Zone, Triangle Visualization, Line Counter

Input and Output Bindings

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

Bindings

• input

  • image (image): not available.
  • detections (Union[instance_segmentation_prediction, object_detection_prediction]): Tracked detections.
  • smoothing_window_size (integer): Predicted movement of detection will be smoothed based on historical measurements of velocity, this parameter controls number of historical measurements taken under account when calculating smoothed velocity. Detections will be removed from generating smoothed predictions if they had been missing for longer than this number of frames..
  • bbox_smoothing_coefficient (float_zero_to_one): Bounding box smoothing coefficient applied when given tracker_id is present on current frame. This parameter must be initialized with value between 0 and 1.

• output

  • tracked_detections (Union[object_detection_prediction, instance_segmentation_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.

Example JSON definition of step Detections Stabilizer in version v1

{
    "name": "<your_step_name_here>",
    "type": "roboflow_core/stabilize_detections@v1",
    "image": "<block_does_not_provide_example>",
    "detections": "$steps.object_detection_model.predictions",
    "smoothing_window_size": 5,
    "bbox_smoothing_coefficient": 0.2
}