Segment Anything 2 Model

Class: SegmentAnything2BlockV1

Source: inference.core.workflows.core_steps.models.foundation.segment_anything2.v1.SegmentAnything2BlockV1

Run Segment Anything 2, a zero-shot instance segmentation model, on an image.

** Dedicated inference server required (GPU recomended) **

You can use pass in boxes/predictions from other models to Segment Anything 2 to use as prompts for the model. If you pass in box detections from another model, the class names of the boxes will be forwarded to the predicted masks. If using the model unprompted, the model will assign integers as class names / ids.

Type identifier

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

Properties

Name	Type	Description	Refs
name	str	Enter a unique identifier for this step..	❌
version	str	Model to be used. One of hiera_large, hiera_small, hiera_tiny, hiera_b_plus.	✅
threshold	float	Threshold for predicted masks scores.	✅
multimask_output	bool	Flag to determine whether to use sam2 internal multimask or single mask mode. For ambiguous prompts setting to True is recomended..	✅

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 Segment Anything 2 Model in version v1.

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

Input and Output Bindings

The available connections depend on its binding kinds. Check what binding kinds Segment Anything 2 Model in version v1 has.

Bindings

• input

  • images (image): The image to infer on..
  • boxes (Union[keypoint_detection_prediction, instance_segmentation_prediction, object_detection_prediction]): Bounding boxes (from another model) to convert to polygons.
  • version (string): Model to be used. One of hiera_large, hiera_small, hiera_tiny, hiera_b_plus.
  • threshold (float): Threshold for predicted masks scores.
  • multimask_output (boolean): Flag to determine whether to use sam2 internal multimask or single mask mode. For ambiguous prompts setting to True is recomended..

• output

  • predictions (instance_segmentation_prediction): Prediction with detected bounding boxes and segmentation masks in form of sv.Detections(...) object.

Example JSON definition of step Segment Anything 2 Model in version v1

{
    "name": "<your_step_name_here>",
    "type": "roboflow_core/segment_anything@v1",
    "images": "$inputs.image",
    "boxes": "$steps.object_detection_model.predictions",
    "version": "hiera_large",
    "threshold": 0.3,
    "multimask_output": true
}